55.2% of the top 10,000 tokens are HIGH RISK. 131 are confirmed honeypots – tokens where you can buy but cannot sell. 1,865 are upgradeable proxy contracts, of which 139 are controlled by a single private key that can silently replace the entire token implementation in one transaction. Only 18.7% pass all 127 security checks and receive a CLEAN verdict.

What Token Audit Found

BNB Smart Chain is the most dangerous chain in the dataset: 68.3% high risk and only 7.6% clean. Ethereum shows 59 confirmed honeypots – tokens that passed as legitimate long enough to enter the CoinGecko top 10,000, then trapped every buyer inside. The two most widespread risk patterns: 35.9% of tokens have no enforceable supply cap (unlimited inflation possible), and 34.4% have no timelock on privileged admin functions (instant malicious governance possible, no delay, no warning).

What Other Tools Miss

Token Audit runs 127 checks across 9 modules – Ownership, Supply, Liquidity, Transfer, Approve, Permit, Pausability, Reentrancy, and Proxy Analysis. The checks that matter most are the ones competitors cannot run: transitive approve() call graph analysis, phantom balanceOf detection, EIP-2612 permit preload, reentrancy invariants, and asymmetric pause detection (pause that blocks sells but not buys). These threats are invisible to GoPlus, CertiK Skynet Token Scan, TokenSniffer, and Honeypot.is – which together cover fewer than 40 checks, all at the interface level. The 599 verdicts in this dataset driven by Approve and Reentrancy findings represent tokens that every competing tool would have passed as clean. This same deployer-behavioral approach powers ChainAware’s token launch auditor agent, which combines contract risk with deployer wallet screening in a single call.

Co-Founder Statement

“We assumed the top 10,000 by market cap would be the safer end of the market. What we found is that more than half carry meaningful risk vectors – and roughly 600 of those are threats that no other automated tool would detect. The sophisticated operators know exactly which checks existing tools run, and they design around them. Token Audit was built to catch what they build.” – Martin Ploom, Co-Founder, ChainAware.ai

Proxy Risk: 139 Tokens One Transaction Away From a Honeypot

Token Audit’s proxy classification goes beyond detecting whether a contract is upgradeable. It identifies who controls the upgrade – and the answer matters enormously. Of 1,865 proxy contracts in the top 10,000, 139 are EOA-controlled: a single unprotected private key can replace the entire implementation in one block, with no timelock, no multisig, no governance vote. BSC carries the highest concentration – 75 of those 139 EOA-controlled proxies are on BSC, where the tactic is a known professional rug pull pattern. On the positive side, 153 UUPS proxies have permanently locked or renounced their upgrade path – Token Audit surfaces this as an explicit positive signal rather than treating all proxies as equally risky. This deployer-level behavioral scoring is the same methodology behind ChainAware’s Wallet Auditor, applied here to contract creators and LP providers instead of individual users.

For the full methodology, chain-by-chain breakdown, finding frequency analysis, honeypot signal correlations, and the complete competitive comparison against GoPlus, TokenSniffer, CertiK Skynet, and Honeypot.is, read the deep-dive: ChainAware Token Audit Launched – We Tested 10,000 CoinGecko Tokens. Here Are the Results. →

ChainAware.ai is the Web3 Agentic Growth Infrastructure – 20M+ wallet personas, 98% fraud detection accuracy, 127-check Token Audit, Agent Trust Score for 274,000+ ERC-8004 agents available via the Prediction MCP. Named in CB Insights’ AI Fraud Prevention Market Map. chainaware.ai

The post 55% of the Top 10,000 CoinGecko Tokens Are High Risk. ChainAware Token Audit Shows Why. first appeared on ChainAware.ai.

The smart contract security audit market is broken. Manual audits cost $5,000 to $150,000 and take weeks. Meanwhile, thousands of new tokens launch every single day – and the vast majority of retail investors check exactly nothing before they buy. On Binance Smart Chain alone, 95% of new liquidity pools end in rug pulls. The tools that exist – GoPlus, TokenSniffer, Honeypot.is – catch the obvious scams. They completely miss the sophisticated ones.

Today, ChainAware is changing that. Token Audit is live: 127 automated security checks across 9 analysis modules, powered by deep code analysis and ChainAware’s behavioral intelligence layer. To validate the system, we ran it against the top 10,000 tokens on CoinGecko, sorted by market capitalization. Those are not random memecoins – they are the most-traded, most-held tokens in crypto. The results are alarming.

This article presents every finding. Specifically, you will learn what the most dangerous patterns look like at scale, which chains produce the highest risk concentrations, and why the tools you are currently using are systematically missing the threats that matter most.

The Study: 10,000 CoinGecko Tokens, 13,000 Audits, 6 Chains

The dataset covers the top 10,000 tokens by market capitalization on CoinGecko as of July 2026. Because many tokens exist simultaneously on multiple blockchains – USDT, for example, runs on Ethereum, BSC, Polygon, Base, and Arbitrum – the total audit count reaches 12,998 individual contract audits across 6 chains. Consequently, each audit is independent: the same token contract deployed on ETH and BSC receives two separate audits, because the contract code, liquidity structure, and ownership configuration can differ significantly between deployments.

Furthermore, this is not a random sample of newly launched tokens. These are established, widely-traded assets – the tokens that appear in your wallet app, on DeFi dashboards, and in portfolio trackers. If findings this severe appear in the top 10,000 by market cap, the situation in the broader universe of hundreds of thousands of tokens is considerably worse.

Each audit runs 127 checks across 9 modules: Ownership, Supply, Liquidity, Transfer, Approve, Permit, Pausability, Reentrancy, and the proprietary Honeypot Pattern module. Three detection layers underpin each audit: deep code analysis for semantic code-level findings, direct on-chain RPC calls for live state verification, and ChainAware’s behavioral database for creator and LP trust scoring. Results are stored in a structured database with one scalar column per finding – enabling the statistical analysis below.

Chain Distribution

Token classification matters for accurate results. Reflection tokens, rebasing tokens, ERC-4626 vault tokens, and bridge tokens all have non-standard transfer mechanics that would trigger false positives in a naive static analysis tool. Token Audit identifies these token types using dedicated classifiers and adjusts its findings accordingly – for example, a reflection token legitimately fails the transfer conservation check by design, and Token Audit documents this distinction rather than incorrectly flagging it as a theft vector. Accurate false-positive management at scale is essential for a tool that will be embedded in high-volume platform integrations, where a false positive on a major legitimate token destroys user trust far faster than a false negative on an obscure scam.

Ethereum leads by audit count, reflecting the concentration of established DeFi protocols on the oldest EVM chain. BSC’s 26.7% share is notable: despite hosting a smaller share of top-10,000 market cap tokens, it accounts for a disproportionate share of the worst findings – as the chain-by-chain breakdown below demonstrates.

The Headline Results: 55% of the Top 10,000 Tokens Are High Risk

The single most important finding from this study is also the most unsettling one. Among the top 10,000 tokens by market cap – the most established, most liquid, most widely held tokens in the entire crypto market – 55.2% receive a HIGH RISK verdict from ChainAware Token Audit.

Only 18.7% of audited tokens receive a CLEAN verdict – meaning they pass all critical security checks, have no meaningful rug pull vectors, and carry no significant code-level risks. Put another way, more than 4 in every 5 tokens in the top 10,000 carry some level of meaningful security concern.

These numbers require context. HIGH RISK does not automatically mean the token is a scam. Many HIGH RISK findings reflect architectural choices that are widespread in legitimate DeFi protocols: uncapped mint functions controlled by governance contracts, upgradeable proxy architectures managed by multisigs, or LP positions not locked because the team chose a different treasury structure. However, HIGH RISK does mean that the token contract contains mechanisms a malicious actor could use to harm investors – and that investors deserve to know about them before committing capital.

Moreover, 131 confirmed honeypots in the top 10,000 is not a small number. These are tokens where the Token Audit’s simulation analysis module confirmed that you can buy – but cannot sell. Twelve of those honeypots were found on Ethereum, the chain most associated with institutional quality and regulatory oversight. The assumption that “top 10,000 by market cap = safe” is demonstrably false.

Results by Chain: BSC Is the Most Dangerous

BSC stands out dramatically. Only 7.6% of BSC token deployments in the top 10,000 are clean – the lowest of any chain in the study. Meanwhile, 68.3% are high risk and another 23.0% are suspicious. Combined, that means 91.3% of top-10,000 BSC tokens carry some security concern. This finding is consistent with BSC’s broader reputation: Chainalysis research identifies BSC as hosting approximately 71% of all rug pull scams globally, driven by lower transaction fees that make deploying fraudulent contracts nearly cost-free.

Base, by contrast, is the cleanest chain in the study at 22.5% clean. Its 40.0% high risk rate reflects a newer, more curated DeFi ecosystem. Nevertheless, 40% high risk across Base’s top tokens is not a reassuring figure.

What Drives the Risk: The Two Dominant Findings

Two findings appear far more frequently than any other in the dataset, together driving 76% of all HIGH RISK verdicts. Understanding them is essential to understanding why so many established tokens carry elevated risk scores.

Finding #1: 35.9% of Tokens Have No Mint Cap (INV_S2_NO_MINT_CAP)

The most common single finding across the entire dataset: 4,668 tokens – 35.9% of all audited contracts – have a mint function with no enforceable supply cap. This means the token’s owner, governance contract, or admin address can create unlimited new tokens at any time, diluting every existing holder’s position to zero.

Critically, this finding appears almost exclusively in HIGH RISK verdicts. Cross-referencing the two columns shows that zero CLEAN tokens carry NO_MINT_CAP – a perfect separation. Every CLEAN token in the dataset either has no mint function at all or has a mint function with an immutable, on-chain cap. The 4,668 NO_MINT_CAP tokens are split between HIGH RISK (4,439) and HONEYPOT (75), with only 154 in the SUSPICIOUS tier.

For investors, the implication is straightforward: a token with an uncapped mint function carries a structural risk that no amount of team credibility or market cap size eliminates. The inflation vector exists regardless of whether the team currently intends to use it.

Finding #2: 34.4% of Tokens Have No Timelock on Privileged Functions (INV_O6_NO_TIMELOCK)

The second-most common finding: 4,470 tokens – 34.4% – have privileged administrative functions (ownership transfer, fee modification, upgrade execution, mint authorization) with no timelock. A timelock requires that any privileged action be announced on-chain and delayed by a minimum period – typically 24 to 72 hours – giving the community time to react if a malicious or compromised admin executes a dangerous change.

Without a timelock, a single administrative transaction can drain a protocol, rug liquidity, or convert a functioning token into a honeypot in a single block. The attacker’s advantage is complete: investors cannot react to changes they cannot anticipate. Adding a timelock costs developers essentially nothing but a few lines of Solidity – which makes its absence in 34.4% of the top-10,000 tokens particularly striking.

Together, NO_MINT_CAP and NO_TIMELOCK account for the overwhelming majority of high-risk verdicts in this dataset. Both findings are invisible to honeypot simulation tools like Honeypot.is – which only checks whether a sell transaction reverts. Furthermore, both are absent from the GoPlus Security API’s detection layer. ChainAware’s Ownership and Supply modules specifically scan for these patterns using deep code analysis, which can trace through function call chains to confirm whether an enforceable cap or delay mechanism actually exists – not merely whether the contract declares one.

Liquidity Risk: 25.7% of Tokens Have Completely Unlocked LP

Beyond the supply and ownership findings, the Liquidity module produced the study’s most operationally urgent results. Liquidity is the primary signal that drives 42% of all verdicts – more than any other module – because liquidity risk is both the most directly dangerous and the most immediately verifiable.

The 25.7% unlocked LP figure is particularly significant. When LP tokens remain in the deployer’s wallet, the entire liquidity backing the token can be removed in a single transaction. Every investor who holds the token is exposed to total loss within one block. The deployer may have committed publicly to never removing liquidity – but without an on-chain lock, that commitment is entirely unenforceable. For how ChainAware detects LP lock status across both V2 (ERC-20 LP tokens) and V3 (NFT positions), see the Liquidity Verification module documentation.

Notably, liquidity lock expiry detection – finding INV_L3_LOCK_EXPIRED – currently has zero hits in the dataset. This finding detects LP locks that have already expired but the associated tokens have not yet been removed. Its absence likely reflects the study’s population: tokens with expired locks often appear after rug pulls have occurred, meaning the token may have been delisted or the pool may have been drained before it entered the CoinGecko top-10,000 dataset.

Confirmed Honeypots: 131 Tokens Where You Can Buy But Cannot Sell

Token Audit’s simulation analysis module forks the relevant blockchain, executes a real buy transaction inside the fork, then attempts a sell. When the sell reverts – meaning the token architecture actively prevents investors from exiting their positions – the verdict is HONEYPOT. This study confirmed 131 honeypots across the top 10,000 CoinGecko tokens.

Ethereum’s 59 confirmed honeypots deserve special attention. The assumption that Ethereum’s higher gas costs and more sophisticated user base filter out honeypot contracts is not supported by this data. Sophisticated honeypots on Ethereum often work precisely because they look legitimate: verified source code, reasonable tax rates, functioning buy mechanics, and professional-looking documentation. The sell block is implemented deep in the transfer call graph – typically using assembly instructions or layered delegation patterns that simple rule-based scanners do not detect.

What Makes a Honeypot: The Three Strongest Signals

Signal 1: hp_CUSTOM_TRANSFER_ENTRY_POINT – Present in 63% of confirmed honeypots (correlation +0.34). This finding fires when the token contract routes transfer calls through a non-standard function before reaching the standard _transfer implementation. Custom entry points are the primary mechanism honeypot developers use to insert sell-blocking logic while keeping the standard ERC-20 interface intact.

Signal 2: hp_UNEXPECTED_EVENTS_IN_TRANSFER – The single highest-correlation honeypot predictor at +0.46, present in 50% of confirmed honeypots. When a transfer function emits events beyond the standard Transfer(from, to, amount) required by ERC-20, it almost always indicates hidden logic inserting itself into the transfer path.

Signal 3: hp_LAYERED_TRANSFER_DELEGATION – Present in 53% of confirmed honeypots and 10.2% of all tokens. Layered delegation means the transfer function calls internal functions that call further internal functions, each potentially adding conditions. Professional honeypots use five or six levels specifically to bury the sell-blocking condition deep enough that automated scanners trace only the outer layers. For how ChainAware’s transfer invariant checking works, see the Transfer Verification documentation.

What Only ChainAware Finds: The Sophisticated Threats

The most significant contribution of this study is not the headline numbers – it is the class of threats that appear in this dataset and cannot be detected by any competing automated tool. ChainAware Token Audit runs 127 checks. Competitors like GoPlus run approximately 40. CertiK Skynet’s free Token Scan runs 19. The gap between those check counts corresponds directly to classes of threat that are invisible to current market-standard tools.

Transfer Conservation Analysis: The Silent Value Drain

Token Audit’s most technically distinctive check is Transfer Conservation (INV_T1_CONSERVATION_FAIL): the invariant that when Alice transfers 100 tokens to Bob, Alice’s balance decreases by exactly 100 and Bob’s balance increases by exactly 100. If sender_lost does not equal recipient_gained, value is being silently diverted – typically to a hidden fee recipient not disclosed anywhere in the token’s interface. Conservation-failing tokens pass every honeypot simulation test. Honeypot.is returns CLEAN. GoPlus returns CLEAN. The investor loses capital on each trade while the token technically allows selling.

The Phantom Balance variant (INV_T5_PHANTOM_BALANCEOF) – found in 5 tokens – is even more sophisticated. The token maintains two separate balance mappings: one that balanceOf() reads and displays to the investor, and a different one that _transfer() actually debits. Your wallet shows you holding 10,000 tokens while the transfer mechanism has already marked your real balance as zero. For the full invariant specification, see the Transfer Invariants documentation.

Permit Correctness: The EIP-2612 Attack Surface

EIP-2612 permit() is implemented in 30% of tokens in this dataset (3,903 tokens). No automated scanner other than ChainAware checks whether the permit implementation is actually correct. Finding INV_P7_PRELOADED_PERMIT – a constructor-time unlimited approval grant – appears in 21 tokens. These 21 tokens allow the deployer to drain any holder’s position at any time using a signature created before any investor bought the token. For how ChainAware detects permit vulnerabilities, see the Permit Verification module.

Approve Security: The Transitive Attack

ChainAware’s Approve module traces the complete call graph of approve() – catching tokens where calling approve(spender, 1000) also silently writes the caller’s balance to zero as a hidden side effect. Finding INV1_EXTRA_STATE_WRITE appears in 63 tokens. INV3_EXTERNAL_CALL_IN_APPROVE appears in 28 tokens. Both require deep code analysis. Neither GoPlus, TokenSniffer, CertiK Skynet, nor De.Fi Scanner runs this analysis. See the Approve Verification documentation.

Reentrancy Analysis

ChainAware is the only automated token scanner that includes reentrancy detection. This study found 540 tokens with no reentrancy guard (INV_R2_NO_REENTRANCY_GUARD), 485 tokens using legacy ETH transfer patterns vulnerable to callback exploitation (INV_R6_ETH_TRANSFER_LEGACY), and 48 tokens with read-only reentrancy exposure (INV_R5_READONLY_REENTRANCY). According to the OWASP Smart Contract Top 10, reentrancy remains one of the most exploited vulnerability categories in DeFi. See the Reentrancy Verification documentation.

Proxy Analysis: 13.2% of Tokens Are Upgradeable Contracts

Token Audit detected proxy contracts in 1,865 tokens – 14.3% of all audited contracts. More importantly, it classifies each proxy by who controls the upgrade function, producing a six-tier risk assessment for the upgrade authority.

Tier	Upgrade Control	Tokens	% of Proxies	Risk
EOA-Controlled	Single private key	139	7.5%	🔴 Critical
Unknown Auth	Cannot be resolved	383	20.5%	🟠 High
Contract-Controlled	DAO / protocol governance	1,152	61.8%	🟡 Medium
Multisig-Controlled	Multiple required signers	29	1.6%	🟢 Low
Timelock-Controlled	Delayed on-chain execution	9	0.5%	🟢 Lowest
UUPS Locked / Renounced	Upgrade permanently disabled	153	8.2%	Immutable

Among all proxy findings, the 139 EOA-controlled proxies represent the most urgent concern. These tokens are upgradeable by a single private key – no multisig, no governance vote, no timelock delay. One transaction from one address can replace the entire contract implementation. BSC accounts for 75 of the 139 EOA-controlled proxies – 54% of the most dangerous proxy tier on less than a third of the token count. For how ChainAware classifies proxy types, see the Ownership Verification module documentation.

Risk Score Analysis: What the Numbers Say at Scale

Liquidity and Supply together drive 76% of all verdicts. The 2.8% driven by Approve and 1.8% by Reentrancy represent high-value findings that no competitor detects. Those 599 verdicts cover the sophisticated operators who invest in clean-looking code specifically to pass GoPlus and TokenSniffer while hiding more subtle attack vectors.

How Token Audit Compares to Existing Tools

Security Check	GoPlus	TokenSniffer	CertiK Skynet	Honeypot.is	ChainAware
Honeypot simulation
Mint capability				❌	+ hidden mint + cap quality
LP lock status			❌	❌	(V2 + V3 NFT positions)
Timelock absence check	❌	❌	❌	❌	Unique
Approve() call graph analysis	❌	❌	❌	❌	Unique
Transfer conservation invariant	❌	❌	❌	❌	Unique
Phantom balanceOf detection	❌	❌	❌	❌	Unique
Permit() correctness (EIP-2612)	❌	❌	❌	❌	Unique
Reentrancy analysis	❌	❌	❌	❌	Unique
Creator behavioral Trust Score	❌	❌	❌	❌	Unique
LP provider Trust Scores	❌	❌	❌	❌	Unique

GoPlus Security is the market standard, averaging 717 million monthly API calls in 2025. Its coverage is broad but rule-based rather than semantic. The approve transitive attack, phantom balance exploit, permit preload, and reentrancy vectors are all invisible to GoPlus’s current architecture. For how ChainAware fits into the broader DeFi security ecosystem, see our Rug Pull Detection Tools comparison and DeFi Compliance Tools guide.

The Behavioral Layer: What Code Analysis Cannot See

Code analysis answers one question: does this contract contain dangerous mechanisms? It cannot answer a more important one: does the person who deployed this contract intend to use those mechanisms maliciously?

This distinction matters because the most dangerous operators specifically invest in clean-looking code. A professional rug pull team in 2026 runs deep code analysis before deploying, checks their own contract against GoPlus, and removes every pattern that produces a red flag. They keep the mint function but make it look like a governance-controlled feature. They leave the LP unlocked but explain it as a treasury management decision. The code passes every automated check. Then, after accumulating enough liquidity, they execute.

ChainAware’s behavioral Trust Score system operates on a fundamentally different signal: the on-chain history of every wallet that deployed the contract and every wallet that provided liquidity. A deployer whose previous contracts ended in rug pulls carries that history regardless of how clean the new contract looks. An LP provider who has removed liquidity from multiple projects within 30 days of launch carries that behavioral signature regardless of how long they have held the current position.

These behavioral signals draw on ChainAware’s core fraud detection infrastructure – the same system that achieves 98% fraud prediction accuracy across 20 million+ wallet behavioral profiles. Combined with the code-level findings from Token Audit’s nine modules, the result is the only token security tool that catches both the technical vulnerability and the operator intent simultaneously. For the full behavioral intelligence methodology, see What Are Web3 Personas and the Fraud Detector documentation.

The Data Moat: Why Token Audit Cannot Be Replicated

Token Audit is built on three proprietary data assets accumulated over years of continuous operation. A competitor starting today cannot purchase these assets, compress the time required to build them, or replicate them from publicly available sources alone. Each one directly enables detection capabilities that require the asset to exist before the analysis can run – meaning the gap between ChainAware and any new entrant widens over time rather than narrowing.

20M+ Wallet Personas: The Behavioral Trust Score Foundation

Every Token Audit includes a creator behavioral Trust Score and LP provider Trust Scores – signals that no competing token scanner offers. These scores draw on ChainAware’s database of more than 20 million wallet behavioral profiles accumulated across 8 blockchains. Each profile represents a complete behavioral fingerprint: transaction history, timing patterns, counterparty networks, protocol diversity, AML exposure, and dozens of derived features trained against confirmed fraud outcomes. The result is 98% fraud prediction accuracy on held-out test data.

This persona depth is what makes the behavioral layer meaningful. A deployer whose previous contracts ended in rug pulls carries that history as a permanent behavioral signal – regardless of how clean the new contract code looks. Without the 20M+ persona database, the behavioral Trust Score would be a near-zero confidence interval. Building that database required years of continuous on-chain data collection and iterative retraining against real-world fraud cases. A new entrant cannot compress that timeline. Furthermore, the model retrains continuously on new confirmed fraud cases – meaning the behavioral edge compounds as ChainAware observes more fraud patterns than any competitor accumulating data from a cold start.

One Year of On-Chain Pair History: The Criminal Record Database

Token Audit’s creator Trust Score cross-references the token deployer’s wallet address against ChainAware’s database of confirmed rug pull and honeypot operators – a database built from more than a year of continuous monitoring of liquidity pair creation and removal events across PancakeSwap, Uniswap, and other major DEX venues. This database records which wallet addresses created pools that subsequently exhibited rug pull patterns, and which wallet addresses previously deployed honeypot token contracts.

This is the data asset that catches the serial scammer deploying a new token after previous campaigns. The rug puller of Q4 2025 is registered as a known criminal in ChainAware’s pair history database. When they deploy a new token in Q1 2026, Token Audit flags the creator wallet immediately – regardless of how clean the new contract code appears. No competitor runs this check because no competitor maintains a paired rug pull database cross-referenced against token deployer wallets. Building it retroactively is also impossible: identifying fraud outcomes requires the passage of time to observe liquidity removal patterns after the fact. The database is a one-year head start that cannot be bought or downloaded. For the data behind this detection layer, see our Rug Pull Tracker report.

Deep Code Analysis Infrastructure: The Semantic Engine Behind 127 Checks

The nine analysis modules that produce Token Audit’s unique findings – approve call graph analysis, transfer conservation invariants, phantom balance detection, permit correctness checking, and reentrancy analysis – all depend on a semantic code analysis infrastructure built specifically for EVM token analysis. It handles Solidity’s inheritance chains, proxy delegation patterns, assembly blocks within Solidity functions, and the non-standard token architectures (reflection, rebasing, ERC-4626 vault tokens) that cause false positives in naive static analysis tools.

Building this infrastructure required years of engineering investment. Every EVM edge case – from DELEGATECALL chains that must be traced across contract boundaries, to assembly-level balance manipulation that bypasses Solidity’s type system, to the layered transfer delegation patterns used by professional honeypot developers – required specific detection logic designed from first principles. The result is a scanner that runs 127 checks in a median of 11.3 seconds across any EVM-compatible contract. That combination of depth and speed is what enables the 9 unique findings in this study that no competitor detects. A new entrant replicating this infrastructure from scratch would need years of engineering time and a corpus of real fraud contracts to validate against – both of which ChainAware has already invested. For the competitive context, see our Forensic vs AI-Powered Blockchain Analysis guide.

Why the Moat Compounds

Each of these three assets improves as it grows. More wallet personas means better fraud prediction precision on creator behavioral scores. More pair history means more confirmed criminal operator wallets in the cross-reference database. More contracts analyzed means more edge cases handled correctly in the deep code analysis infrastructure. A competitor starting today with identical engineering resources would still need years to reach ChainAware’s current capability level – and by then, ChainAware’s data assets would be proportionally larger still. The advantage is a trajectory, not a snapshot.

What Does a CLEAN Token Look Like?

2,436 tokens in this dataset – 18.7% – received a CLEAN verdict. Understanding what they have in common is as instructive as understanding what HIGH RISK tokens share.

Notably, zero CLEAN tokens have an uncapped mint function. Every CLEAN token either has no mint capability at all, or has a mint function with an immutable, verifiable on-chain cap. This single characteristic is the strongest predictor of a clean verdict – more consistent than any other single check in the dataset.

Additionally, CLEAN tokens overwhelmingly have verified source code. Their LP is either locked in a recognized locker (PinkLock, UniCrypt, Team Finance), burned to a dead address, or the project has explicitly structured treasury management differently with transparent on-chain documentation. Their ownership model is either renounced, controlled by a multisig with public signers, or timelocked. The transfer function has no assembly in its call graph, no external calls, and emits exactly the events ERC-20 requires – nothing more, nothing less. For the complete CLEAN verdict criteria, see the Token Audit Verdict Methodology.

Implications for Investors, Platforms, and Builders

For Individual Investors

The core finding of this study is that market capitalization rank is not a security signal. Tokens in the CoinGecko top 10,000 are 55.2% high risk and 1% confirmed honeypot. Before committing capital to any token, check three things specifically: whether the LP is locked, whether the mint function has an enforceable cap, and whether the contract is upgradeable by a single EOA. ChainAware Token Audit checks all three – and 124 other things – in under 60 seconds, free, without requiring a wallet connection. For how to interpret Token Audit results, see the Token Audit Investor Guide.

For DeFi Platforms and DEX Aggregators

Platforms that surface token information currently rely almost entirely on GoPlus for token security data. This study demonstrates that GoPlus-equivalent analysis leaves substantial risk categories completely undetected. Embedding Token Audit results at the listing or interaction point gives users substantially more protection than any current alternative. The REST API and MCP integration return full structured results including per-finding boolean flags, per-module risk scores, and a human-readable verdict. For technical integration details, see the Token Audit API documentation and the MCP Integration guide.

For Token Builders

The 18.7% CLEAN rate in this study is not a verdict on intent – most high-risk findings reflect architectural patterns that developers adopted without understanding their security implications. Token Audit runs in full against any deployed contract, returning specific findings with remediation guidance for each. Running Token Audit costs nothing and takes 60 seconds. It identifies every architectural risk that investors, security researchers, and automated tools will find after deployment – and gives developers the opportunity to fix them first. For how to use Token Audit in a pre-deployment security review, see the Pre-Deployment Checklist.

Pausability: 5% of Tokens Can Freeze All Trading Right Now

This study found 648 tokens – 5.0% of all audited contracts – where the mint function continues operating even when the token is paused (INV_PA5_MINT_NOT_PAUSED). An admin can pause all investor transfers while continuing to mint new tokens into their own wallet – simultaneously trapping existing holders and diluting their positions.

The most sophisticated pausability vulnerability – INV_PA4_ASYMMETRIC_PAUSE – blocks sells (transferFrom) while allowing buys (transfer). Honeypot.is tests a sell by calling transfer – the same function that is allowed in the asymmetric pause scenario – so it returns CLEAN for a token that is functionally a honeypot. ChainAware detects the asymmetric pattern by analyzing whether the pause condition applies differently to transfer versus transferFrom.

The 22 tokens currently paused represent an immediate alert for any investor holding these tokens. Token Audit calls paused() directly on each pausable contract to determine whether the pause is currently active – those 22 tokens are actively frozen right now. See the Pausability Verification documentation.

Supply Analysis: Hidden Minting and Supply Manipulation at Scale

Finding INV_S1_HIDDEN_MINT appears in 822 tokens – 6.3% of the dataset. Hidden mint detects functions that inflate the total token supply through mechanisms not labeled as mint() or _mint(). Because they bypass the standard _mint internal function, simple checks that scan for mint selectors in the contract ABI will miss them entirely. ChainAware traces every function that modifies the total supply variable regardless of name. See the Supply Verification documentation.

Finding INV_S4_FAKE_BURN appears in 318 tokens – 2.4%. A fake burn transfers to address(0) but does not reduce totalSupply(). Tokens marketed as deflationary based on burn history may be inflating their apparent scarcity. Additionally, 216 tokens show deployer concentration at 100% of circulating supply (INV_S6_DEPLOYER_100PCT) – the optimal setup for a coordinated pump-and-dump.

Audit Performance: 15 Seconds Average, 98.4% Source Verified

The median audit completes in 11.3 seconds – well within the threshold for interactive use cases like a DEX listing flow or a wallet pre-transaction security check. Only 9 audits across the entire 12,998-audit dataset exceeded 120 seconds – representing 0.07% of cases and well within operational tolerances for any integration scenario. 98.4% of tokens in this dataset have verified source code. For unverified contracts, Token Audit operates in bytecode analysis mode – the Honeypot Pattern module, Liquidity module, Simulation module, and Behavioral Trust Score all operate on bytecode and on-chain state rather than source code. For details see the Unverified Contract Analysis documentation.

Conclusion: The Token Security Gap Is Real

This study set out to answer a simple question: how safe are the tokens that most investors actually hold? The answer – 55.2% high risk, 1% confirmed honeypot, 13.2% upgradeable proxy, 25.7% unlocked LP – is more alarming than most observers expected from the top 10,000 by market capitalization. These are not obscure tokens in forgotten DEX pools. Many appear in mainstream wallet apps, on regulated exchange listings, and in institutional portfolio allocations.

Furthermore, the findings that established tools miss are precisely the ones that matter most for sophisticated attacks. GoPlus, TokenSniffer, and CertiK Skynet catch the obvious patterns. Consequently, professional scam operators have adapted: they write clean-looking code that passes all three tools, then execute through vectors those tools cannot see. The approve transitive attack, the phantom balance exploit, the permit preload, the asymmetric pause – all of these appear in this dataset, and all of them are invisible to current market-standard scanners.

ChainAware Token Audit changes this equation. It brings institutional-grade deep code analysis to every token, automatically, for free, in under 60 seconds. Combined with simulation analysis, behavioral Trust Scores, and proxy upgrade authority classification, Token Audit produces a security profile that exceeds what any competing automated tool provides. According to FATF’s Virtual Assets Recommendations, real-time token screening is becoming a compliance requirement for virtual asset service providers globally. Token Audit is live today – test any contract free, no signup, no wallet connection. For enterprise integration, book a technical walkthrough below.

Frequently Asked Questions

What is a token audit?

A token audit is an automated or manual security review of a cryptocurrency token’s smart contract. A manual token audit performed by firms like CertiK or Hacken costs $5,000 to $150,000 and takes one to four weeks. ChainAware Token Audit performs automated analysis across 127 checks in under 60 seconds at no cost for individual queries.

How is ChainAware Token Audit different from GoPlus?

GoPlus Security runs approximately 717 million monthly API calls. Its detection is rule-based – it checks for known dangerous patterns at the interface level. ChainAware Token Audit adds semantic analysis via deep code analysis, which traces the complete execution paths of transfer(), approve(), and related functions to find vulnerabilities hidden deep in internal call chains. ChainAware also adds reentrancy detection, permit correctness analysis, supply consistency checking, and behavioral Trust Scores – none of which GoPlus offers.

What does HIGH RISK mean in practice?

HIGH RISK means the token contract contains one or more mechanisms that a malicious or compromised admin could use to harm investors – an uncapped mint function, unlocked LP, EOA-controlled proxy, or admin with no timelock. HIGH RISK does not mean the token is currently being exploited – it means the architectural risk exists and investors should evaluate it consciously before committing capital.

How does the simulation analysis work?

Token Audit’s simulation module forks the relevant blockchain at the current block using an Anvil instance, then executes real buy and sell transactions inside the fork. This catches dynamic honeypot behavior that static analysis cannot detect: tokens where sells revert, tokens where the effective sell tax differs from the declared sell tax, and tokens where token conservation fails. See Simulation Module documentation.

Which chains does Token Audit cover?

Token Audit currently supports Ethereum, BNB Smart Chain, Base, Polygon, and Arbitrum. Optimism support is in progress. The analysis architecture is chain-agnostic at the contract level: deep code analysis, ownership tracing, and supply verification work identically across EVM-compatible chains.

What does the creator behavioral Trust Score measure?

The creator Trust Score evaluates the on-chain behavioral history of the wallet that deployed the token contract. It draws on ChainAware’s database of 20 million+ wallet behavioral profiles to assess whether the deployer has patterns consistent with fraud operators – prior rug pulls, coordination with known scam wallet clusters, funding source characteristics, and behavioral sequences associated with professional exit scam operations. See Fraud Detector documentation.

Can Token Audit be used pre-deployment?

Token Audit requires a deployed mainnet contract address – it analyzes live on-chain state alongside contract code. For pre-deployment security review, ChainAware recommends running deep code analysis directly against the contract source, then running Token Audit immediately after mainnet deployment. According to the Smart Contract Weakness Classification Registry, the majority of token vulnerabilities are deterministic at the code level and identifiable through static analysis shortly after deployment.

Sources: Chainalysis Crypto Crime Report · OWASP Smart Contract Top 10 · Smart Contract Weakness Classification Registry · EIP-2612: Permit Extension for ERC-20

Related ChainAware Reading: Best Rug Pull Detection Tools 2026 · DeFi Compliance Tools Comparison · KYT and AML Guide for DeFi · Web3 Wallet Auditing Providers 2026 · What Are Web3 Personas · Prediction MCP for AI Agents · The Web3 Agentic Economy · Agent Trust Score: On-Chain Trust Scoring for ERC-8004

The post ChainAware Token Audit Launched – We Tested 10,000 CoinGecko Tokens. Here Are the Results. first appeared on ChainAware.ai.

103,695 separate rug pull events. $1,947,176,810 in liquidity removed by contract creators. $1,377,788,426 added before removal. The difference – $569,388,384 – flowed directly out of retail wallets and into the pockets of fraudulent actors operating industrial-scale extraction infrastructure on one of the world’s largest decentralized exchanges. Every single week. Quietly. Without ceremony.

This is the rug pull industry. Not a bug. An industry.

Today, ChainAware.ai publishes this data for the first time – and simultaneously launches Rug Pull Detector V3, our most advanced version yet, with 90.1% prediction accuracy achieved by combining behavioral analysis of contract creators with full smart contract inspection. This is the complete picture: what the data shows, why nobody talks about it, how the extraction machinery works, and how to stop it.

In This Guide

• The $569M Dataset: What We Measured and How
• Week-by-Week Breakdown: 20 Weeks of Retail Extraction
• The Industry Silence Problem: Why Nobody Talks About Rug Pulls
• How Rug Pulls Work: The Mechanics of Liquidity Extraction
• Beyond Basic Rug Pulls: The More Complex Extraction Methods We Did Not Count
• Rug Pull Detector V3: From 68% to 90.1% Prediction Power
• How the V3 Algorithm Works: Behavioral + Smart Contract Analysis
• Algorithm Verification and Accuracy Methodology
• Who Uses Rug Pull Detector: Retail Investors, Businesses, and AI Agents
• Projection: How Many Rug Pulls in the Next 20 Weeks?
• The Complete Protection Stack for DApps and Retail Investors
• Frequently Asked Questions

The $569M Dataset: What We Measured and How

ChainAware analyzed every liquidity event on PancakeSwap V2 across weeks 1 through 20 of 2026. The methodology is deliberately conservative. We measured only the most basic, unambiguous form of rug pull: a contract creator adds liquidity to a pool, then removes more than they added. The difference between liquidity added and liquidity removed – when removal exceeds addition – constitutes the rug pull value we report.

This definition intentionally excludes more sophisticated extraction methods. Complex multi-step schemes involving LP token transfers, associated wallet sell-offs, and unlocked token dumps are not included in these numbers. The $569M figure represents the floor – the minimum provably fraudulent extraction we could measure with mathematical certainty from on-chain data alone.

The full dataset covers:

• Total rug pull events detected: 103,695
• Total liquidity added by creators (Mints): $1,377,788,426
• Total liquidity removed by creators (Burns): $1,947,176,810
• Net extraction (Burns minus Mints): $569,388,384
• Period: Week 1 through Week 20, 2026
• Exchange: PancakeSwap V2 (BNB Chain)

PancakeSwap V2 on BNB Chain is one of the highest-volume decentralized exchanges in the world. It is also, by our measurement, one of the largest venues for systematic retail investor fraud. The combination of low gas fees, high token creation velocity, and large retail liquidity makes BNB Chain the preferred operating environment for industrial-scale rug pull operations.

For context on just how large this ecosystem of fraud is: the Bybit hack in February 2025 – which generated enormous industry coverage, emergency response coordination, and weeks of Twitter discussion – extracted $1.46 billion. Our 20-week PancakeSwap V2 measurement represents 39% of that headline-grabbing figure. On one DEX. In one 20-week window. With virtually zero media coverage.

Week-by-Week Breakdown: 20 Weeks of Retail Extraction

The weekly data reveals patterns that any serious analyst of DeFi fraud should study carefully. Rug pull activity is not random noise. It follows detectable rhythms tied to market sentiment, BNB price movements, and the operational cadence of the fraud factories running these schemes.

What the Weekly Patterns Tell Us

Four distinct patterns emerge from careful analysis of this 20-week dataset. Each pattern carries significant implications for how rug pull operations are structured and how retail investors can protect themselves.

Pattern 1 – Early surge, then stabilization: Weeks 1 through 5 represent the highest-intensity period, with Week 4 peaking at $53.4M in a single week. This early-year surge correlates with January-February market optimism, when retail capital flows into DeFi most aggressively following holiday periods. Fraud operators know this and deploy capital accordingly.

Pattern 2 – Persistent baseline fraud: Even in quieter weeks (W8, W17, W18, W19), fraud never drops to zero. The range of $12.6M to $16.7M represents what we call the baseline rug pull floor – the irreducible minimum level of extraction that persists regardless of market conditions. These weeks demonstrate that rug pulls are not opportunistic responses to bull markets. They are a continuous, professionally operated business.

Pattern 3 – Pool creation velocity diverging from fraud events: Week 2 shows the highest pool creation (24,145 total pools) but does not produce the highest fraud value. Week 4 shows far fewer pools (17,984) but delivers the peak fraud extraction. This divergence suggests that fraud operators optimize for pool value, not pool volume – they are creating fewer but more lucrative pools rather than flooding the market with low-value scam tokens.

Pattern 4 – Week 20 resurgence: After a six-week decline from W14 through W19, Week 20 spikes back to $39.9M – the second-highest single-week figure in the dataset. This resurgence suggests cyclical fraud campaigns that compress activity during low-sentiment periods and re-accelerate when retail interest returns. The Week 20 data point is a leading indicator, not an anomaly.

The Industry Silence Problem: Why Nobody Talks About Rug Pulls

Here is a question worth sitting with: Why does a $1.46B exchange hack generate hundreds of articles, emergency response calls, and weeks of industry-wide discussion – while $569M in retail losses across 20 weeks generates almost no coverage at all?

The answer is structural. Exchange hacks are dramatic, concentrated, and attributable. They happen to institutions – Bybit, Binance, Euler Finance – that have PR teams, legal counsel, and market presence. Rug pulls happen to anonymous retail investors, dispersed across thousands of small transactions, with no single victim large enough to command media attention and no single perpetrator identifiable enough to pursue.

$53.4M stolen from a major exchange in a single transaction is a crisis. $53.4M stolen from 6,000 retail investors across thousands of micro-transactions in a single week is business as usual. The victims are too distributed to organize. The fraud is too normalized to shock. The perpetrators are too anonymous to name.

This structural invisibility benefits the rug pull industry enormously. Unlike hacks – which trigger security audits, insurance claims, regulatory investigations, and protocol upgrades – rug pulls operate in a consequence-free zone. The operators who ran this week’s $28M extraction will run next week’s extraction with identical infrastructure. Nothing stops them. Nobody is watching. The industry is busy discussing the latest Layer 2 throughput numbers.

The Chainalysis 2025 Crypto Crime Report documents $17 billion in total crypto scam losses – a figure that includes rug pulls, phishing, fake investments, and other fraud categories. Rug pulls constitute one of the largest subcategories within this total, yet they receive proportionally far less analytical attention than flash loan exploits, bridge hacks, and protocol vulnerabilities.

Part of the problem is that rug pulls do not fit the traditional Web3 security narrative. Security firms make revenue from smart contract audits, not from behavioral fraud detection. Media outlets generate traffic from dramatic hacks, not from steady-state extraction statistics. Influencers amplify projects that pay them, not warnings that might upset project teams with marketing budgets. The incentive structure of the crypto information ecosystem systematically underweights the most consistent form of retail harm.

ChainAware exists, in part, to change this. Publishing this data is the first step. Building tools that let retail investors and DApps verify tokens and pools before committing capital is the second. For a broader perspective on the comparison between rug pulls and pump-and-dump schemes and how each extracts value from retail investors, our dedicated guide covers the full mechanics of both fraud types.

How Rug Pulls Work: The Mechanics of Liquidity Extraction

Understanding rug pull mechanics is the foundation of avoiding them. The basic liquidity rug pull – the type we measured in this dataset – follows a repeatable, five-step operational sequence that any investor can learn to recognize.

Step 1: Token and Pool Creation

A fraudulent actor deploys a new ERC-20 or BEP-20 token contract. On BNB Chain, this costs a fraction of a dollar and takes minutes. The token contract typically includes mint functions (allowing the creator to generate unlimited supply), hidden transfer restrictions (preventing buyers from selling), and ownership functions that are often not renounced. The pool is created on PancakeSwap V2 by pairing the new token with BNB or USDT, establishing an initial price.

Step 2: Liquidity Seeding

The creator deposits initial liquidity – real BNB or USDT – into the pool alongside the worthless token they created. This establishes a tradeable pair and makes the token appear legitimate on DEX aggregators and portfolio trackers. The liquidity seed is the bait. It represents the money the fraudster is temporarily risking to attract retail capital.

Step 3: Marketing and Social Momentum

Telegram groups are created. Twitter accounts post about the new token. Sometimes influencers are paid to promote it. Bots generate artificial trading volume to make the token appear active. Retail investors see a token with real liquidity, active trading, and social proof – and buy in. Each purchase increases the price and deepens the liquidity pool with real capital.

Step 4: Liquidity Removal

Once sufficient retail capital has entered the pool, the creator burns their LP tokens – withdrawing all liquidity from the pool. This single transaction extracts all the BNB or USDT that retail investors deposited when they bought the token. The token price drops to zero instantly. Holders are left with worthless tokens that cannot be sold. The creator walks away with the full liquidity amount, minus their initial seed deposit. The difference is pure profit extracted from retail investors.

Step 5: Repeat

Professional rug pull operators do not run one scheme. They run dozens simultaneously, across multiple wallets, with industrialized tooling that automates contract deployment, social media posting, and liquidity management. The 103,695 rug pull events in our dataset represent the output of a mature industry with operational infrastructure comparable to a sophisticated affiliate marketing operation – except the product being sold is fraud.

Beyond Basic Rug Pulls: The More Complex Extraction Methods We Did Not Count

Our $569M figure covers only the most conservative, mathematically verifiable form of rug pull: direct liquidity removal exceeding direct liquidity addition, by the same contract creator. This definition was chosen deliberately – it produces numbers that cannot be disputed, because they are derived purely from on-chain transaction data with no inferential assumptions.

However, professional rug pull operators frequently use more sophisticated extraction methods that our current measurement excludes. Understanding these methods matters for two reasons: first, because they represent additional losses not captured in the $569M figure; second, because ChainAware’s Rug Pull Detector V3 is being extended to detect these more complex patterns in future iterations.

LP Token Transfer Rug Pulls

Instead of burning LP tokens directly, the creator transfers them to a secondary wallet before burning. This adds one intermediary step that breaks the direct creator-to-burn attribution our basic methodology requires. The economic outcome is identical – retail liquidity is extracted – but the attribution chain is one step longer. Detecting this pattern requires tracking LP token ownership across wallets, not just monitoring mint/burn events on the original creator address.

Unlocked Token Sell-Offs

Some rug pulls never involve liquidity removal at all. Instead, the creator holds a large pre-minted supply of unlocked tokens and sells them gradually into the market as retail buyers push the price up. This is the “slow rug” – a controlled sell-off that extracts value over days or weeks rather than in a single transaction. Detecting this requires monitoring creator wallet sell behavior relative to price action, not liquidity pool events.

Associated Party Extraction

Sophisticated operators fund multiple secondary wallets that receive token allocations at launch and sell into retail buying pressure. The creator’s primary wallet never transacts after the initial liquidity seed – only associated wallets do. Connecting these wallets to the creator requires graph analysis of funding transactions, not just monitoring of the deployer address.

Honeypot Contracts

A honeypot is a contract that allows buying but blocks selling. Transfer restrictions embedded in the contract – hidden from Etherscan views unless you know where to look – prevent token holders from executing sell transactions. Buyers accumulate tokens they cannot sell, while the creator sells their pre-minted allocation freely. GoPlus Security detected 67,241 honeypot tokens in Q4 2024 alone – a figure that underscores the scale of this specific fraud variant. For a full comparison of rug pull detection tools that cover honeypot analysis, see our guide to best Web3 rug pull detection tools in 2026.

The conservative $569M figure – our confirmed minimum – would be substantially higher if all these additional extraction methods were included. ChainAware’s V3 algorithm already incorporates smart contract analysis that can detect honeypot patterns and some forms of unlocked token sell-off risk. Future iterations will extend this coverage further.

Rug Pull Detector V3: From 68% to 90.1% Prediction Power

The previous version of ChainAware’s Rug Pull Detector operated at approximately 68% prediction accuracy. For retail investors, that accuracy level is better than nothing – significantly better than the zero-analysis approach most investors use. However, 68% means that roughly one in three high-risk pools would pass the detector without a warning, and some legitimate pools would trigger false positives.

Our customers asked directly: can we do better? The answer is V3 – and the answer is yes.

ChainAware Rug Pull Detector V3 achieves 90.1% prediction accuracy. This jump from 68% to 90.1% represents a 32.5% relative improvement in prediction power – the largest single-version upgrade in the detector’s history. The improvement comes from a fundamental architecture change: V2 relied exclusively on behavioral analysis of contract creators. V3 combines behavioral analysis with full smart contract inspection.

Why the V2 Accuracy Gap Existed

Behavioral analysis alone – examining the on-chain history of the wallet that deployed a contract – is a powerful signal, but it has a ceiling. Experienced fraud operators know that behavioral signals can be spoofed. They create fresh wallets with clean histories. They fund deployer wallets through legitimate channels. They space out deployments to avoid clustering signals that behavioral models flag.

A behavioral model trained purely on creator wallet history will inevitably miss sophisticated operators who invest in maintaining clean deployer identities. This is the category that the 32% gap in V2 accuracy primarily represented. The false negatives were concentrated in professional, well-organized fraud operations – precisely the operators responsible for the largest individual rug pull events.

What V3 Adds: Smart Contract Analysis

Smart contract analysis reads the code of the contract itself – not just the history of the wallet that deployed it. Regardless of how clean a deployer wallet’s history looks, a contract with hidden mint functions, owner-only transfer restrictions, or unchecked liquidity lock mechanisms will trigger V3’s contract analysis layer.

This combination closes the gap that sophisticated fraud operators had exploited in V2. A fraudster who maintains a clean wallet but deploys a honeypot contract now triggers the smart contract analysis layer even when the behavioral analysis layer returns a clean signal. Conversely, a wallet with minor behavioral flags but a fully transparent, auditable contract receives a more accurate risk assessment that prevents false positives on legitimate projects.

The 90.1% accuracy figure represents the combined performance of both layers together – it is the V3 ensemble model’s prediction power, not either layer in isolation. The algorithm remains under active development. We expect accuracy to continue improving as the training dataset expands and additional smart contract analysis patterns are incorporated.

How the V3 Algorithm Works: Behavioral + Smart Contract Analysis

V3 runs two parallel analysis pipelines that produce independent risk scores, then combines them through an ensemble model trained on verified historical rug pull events. Both pipelines run in real time – the full analysis completes in under two seconds for any pool or token address submitted to the detector.

Pipeline 1: Creator Behavioral Analysis

The behavioral analysis pipeline examines the complete on-chain history of the wallet that deployed the contract. ChainAware’s 20M+ wallet persona database, trained across 8 blockchains, provides the foundation for this analysis. The pipeline evaluates multiple behavioral dimensions that collectively predict fraudulent intent:

• Deployment history: How many contracts has this wallet deployed? What is the historical fate of those contracts – did their pools maintain liquidity or were they rugged?
• Funding provenance: Where did the capital to seed liquidity originate? Wallets funded from known mixer outputs, fresh exchange withdrawals, or clusters of associated addresses receive elevated risk scores.
• Creator feeder analysis: Wallets that funded the deployer are also examined. A deployer wallet with a clean history but funded by a wallet with prior rug pull associations triggers a feeder-chain risk signal.
• Temporal patterns: How quickly after token deployment was liquidity removed in prior contracts from this wallet or associated wallets? Short hold periods are a strong predictor of rug pull intent.
• Wallet age and diversity: Fresh wallets with minimal on-chain history and a single purpose (token deployment and liquidity management) score significantly higher than wallets with years of diverse on-chain activity.

This behavioral layer is powered by the same predictive intelligence that drives ChainAware’s broader wallet analysis capabilities. For context on how the underlying wallet behavioral analysis works across other use cases, our guide to using the ChainAware Wallet Auditor covers the full 9-parameter profile in detail. The full methodology behind this deployer-level scoring is also documented on the Rug Pull Detector learn page.

Pipeline 2: Smart Contract Analysis

The smart contract analysis pipeline inspects the deployed contract code directly. For verified contracts (where source code is published), the analysis performs AST (Abstract Syntax Tree) parsing – examining the structural logic of the contract to identify dangerous patterns. For unverified contracts, bytecode inspection is used to detect characteristic opcode sequences associated with honeypot restrictions and hidden mint functions.

The contract analysis examines specific risk patterns:

• Hidden transfer restrictions: Functions that block selling by non-owner addresses, often disguised within complex conditional logic that is not obvious from casual code review.
• Owner-privileged mint functions: Unrestricted mint capabilities controlled by the deployer allow infinite token supply expansion after retail investors have bought in.
• Ownership renouncement status: Contracts that have not renounced ownership retain the ability to modify transfer restrictions, fee structures, and other parameters after launch.
• Liquidity lock verification: Whether LP tokens are locked – and in what contract, with what unlock conditions – is a critical signal. Unlocked LP tokens in the deployer’s wallet represent immediate rug pull risk.
• Fee manipulation functions: Contracts with owner-callable functions to increase buy/sell taxes after launch can effectively trap investors by making selling economically unviable.

For additional context on how AI-powered smart contract analysis compares to traditional audit approaches, our guide to AI-powered blockchain analysis and machine learning for crypto security covers the broader landscape of ML-based fraud detection.

The Ensemble Model

Outputs from both pipelines feed into an ensemble model that produces a single composite risk score between 0 and 100. Scores above 75 trigger a high-risk warning. Scores between 50 and 75 generate a medium-risk flag with specific risk factors highlighted. Scores below 50 return a lower-risk assessment – though not a guarantee of legitimacy, since the algorithm continues to develop and some novel fraud patterns may not yet be captured.

The ensemble model is trained on a labeled dataset of confirmed rug pull events – including events from the 103,695 cases in our PancakeSwap V2 analysis – and updated continuously as new rug pull events are confirmed. This continuous retraining is what allows the algorithm to adapt to evolving fraud operator tactics rather than becoming outdated as the fraud industry develops new evasion methods.

Algorithm Verification and Accuracy Methodology

The 90.1% accuracy figure requires explanation of its methodology, because prediction accuracy claims in the crypto security space are frequently made without rigorous verification frameworks. ChainAware’s accuracy is measured against historical confirmed rug pull events using a held-out test set – contracts and pools that were not included in the training data but whose eventual rug pull status is now known.

Full verification methodology, including the test set composition, evaluation metrics, false positive and false negative rates by pool type, and comparison to V2 baseline performance, is published at chainaware.ai/resources/rugpull-verification ↗.

Three important caveats about the 90.1% figure deserve explicit statement:

Caveat 1 – The algorithm is in active development. 90.1% is the current measured performance. We expect this number to improve. We also acknowledge that novel fraud patterns not yet present in our training data could temporarily reduce real-world performance below this benchmark.

Caveat 2 – False negatives exist. Roughly 9.9% of rug pull events will not be flagged by V3. These are concentrated in the most sophisticated fraud operations – those that maintain clean creator wallets and deploy contracts that pass automated inspection. Human review of tokenomics, social channels, and team identity remains important for high-value investment decisions.

Caveat 3 – False positives also exist. Some legitimate projects will receive elevated risk scores, particularly those deployed by newer wallets without extensive on-chain history or those using novel contract architectures that match patterns we associate with fraud. V3 is designed to flag risk, not to definitively declare fraud – the final investment decision always rests with the human investor.

For comparison with how other detection tools benchmark their accuracy, the broader context of the forensic versus AI-powered blockchain analysis framework explains why predictive accuracy figures differ fundamentally from forensic identification rates.

Who Uses Rug Pull Detector: Retail Investors, Businesses, and AI Agents

ChainAware’s Rug Pull Detector V3 serves three distinct user categories, each with different integration paths and use case requirements.

Retail Investors: Free Web Tool

Individual investors access the Rug Pull Detector at no cost through the web interface at chainaware.ai/rugpull. No account creation is required. Submit any pool address or token contract address on supported chains, and V3 returns a complete risk analysis within seconds. The tool handles both liquidity pool addresses (where additional LP-specific checks run) and regular token contract addresses.

For retail investors who want to go beyond rug pull risk and understand the complete behavioral profile of any wallet – including their own – the ChainAware Wallet Auditor provides a 9-parameter profile covering experience, risk willingness, intentions, AML status, and Wallet Rank. Both tools are free and require no signup. The full profile methodology is documented on the Wallet Auditor learn page.

DApps and Businesses: API Subscription

Businesses that need to check new pools or existing tokens at scale – DEX aggregators, portfolio trackers, launchpads, DeFi protocols – can subscribe to the Rug Pull Detector API. The API provides the same V3 analysis through a programmatic interface, enabling integration into existing DApp infrastructure without requiring users to manually submit addresses.

For DApps that want to screen wallet connections in real time – catching bad actors at the moment of wallet connection rather than at the moment of investment – the broader ChainAware suite integrates via Google Tag Manager pixel with zero code changes. See the complete Transaction Monitoring Agent guide for how automatic wallet screening works in a live DApp environment.

DApps that want to understand not just fraud risk but the full behavioral profile of their user base – their intentions, experience levels, and conversion likelihood – can combine the Rug Pull Detector API with ChainAware’s Web3 Behavioral User Analytics for a complete picture of who is using their platform and which users represent fraud risk versus conversion opportunity.

AI Agents: X402 Micropayment Protocol

AI agents operating autonomously in DeFi environments – executing trades, managing portfolios, or conducting due diligence on behalf of human principals – can access Rug Pull Detector V3 through the X402 micropayment protocol. This enables agents to pay per analysis in real time without requiring pre-approved API keys or subscription agreements.

An agent evaluating whether to provide liquidity to a new pool, or whether to purchase a newly launched token as part of a portfolio strategy, can query V3 and receive a risk assessment as part of its decision-making pipeline. This integration pattern – AI agents using on-chain behavioral intelligence to make better decisions – is the core use case that ChainAware’s MCP integration guide covers in detail, and the full setup process is documented on the Prediction MCP learn page. For the complete framework of how AI agents are replacing human functions in Web3, our guide to the Web3 agentic economy provides essential context.

Projection: How Many Rug Pulls in the Next 20 Weeks?

Based on the 20-week dataset, what does the next 20-week period look like? This is not a rhetorical question. The data provides enough signal to make a structured projection.

The Baseline Projection

Averaging the 20 weeks of data gives a mean weekly rug pull extraction of approximately $28.5M (total $569.4M divided by 20 weeks). If the next 20 weeks perform at the historical mean, the projected extraction would be approximately $570M – virtually identical to the first 20-week period.

However, averages obscure the variance that makes this projection more nuanced. Week 17 produced just $12.6M. Week 4 produced $53.4M. The range is wide – and what drives the range matters for any forward projection.

Bull Market Scenario

If BNB and the broader crypto market experience a significant price rally in the next 20 weeks, retail capital inflows to PancakeSwap V2 will increase. More retail capital entering DEX pools means more capital available for extraction. In a bull market scenario, weekly extraction figures could return to the W3-W5 range ($38M-$53M per week), pushing the 20-week total toward $800M-$900M.

Bear Market Scenario

Sustained market decline reduces retail participation in speculative DeFi activity. The W16-W19 pattern – four consecutive low weeks averaging $15M – represents what a low-sentiment environment looks like. In a bear market scenario, the next 20-week total could fall to $250M-$350M.

The Conservative Certainty

Regardless of market direction, one conclusion is near-certain: rug pulls will continue to extract hundreds of millions of dollars from retail investors in the next 20 weeks. The fraud infrastructure exists, is profitable, and faces no meaningful deterrent. The floor established by the W17-W19 data ($12.6M-$15.1M per week) implies a minimum 20-week extraction of approximately $250M under the most pessimistic scenario.

The only variable the industry controls is how many investors check tokens and pools before buying – and how accurate those checks are. That is the market for Rug Pull Detector V3.

The Complete Protection Stack for DApps and Retail Investors

Rug pull detection is one layer of a complete Web3 fraud protection stack. Understanding where it sits relative to other security tools helps both retail investors and DApp teams build comprehensive protection rather than relying on any single tool.

For Retail Investors: The Three-Check Protocol

Before committing capital to any new token or pool, a thorough retail investor runs three checks. First, they use ChainAware’s Rug Pull Detector to assess the pool or token’s fraud risk – covering both creator behavior and contract analysis. Second, if the rug pull check flags concerns, they use the Fraud Detector to drill into the specific wallets associated with the token deployment. Third, before sending funds to any individual wallet address, they use the Wallet Auditor to assess the receiving address’s behavioral profile and AML status.

This three-check approach takes less than five minutes per investment decision and provides protection against the most common forms of DeFi fraud. It will not catch every sophisticated attack – nothing will – but it filters out the vast majority of the 103,695 rug pull events in our dataset, which are predominantly straightforward enough to be detected by V3’s behavioral and contract analysis.

For a complete understanding of the crypto security landscape and how different tools protect against different threat vectors, the guide to crypto wallet security in 2026 covers hardware wallets, behavioral intelligence, and fraud prevention in a single comprehensive framework.

For DApps: Pre-Connection Screening

DApps face a different version of the rug pull problem: they are not the ones buying potentially fraudulent tokens, but they are being used as distribution channels by users who may have funded their activity through fraud proceeds. A DApp that allows a rug pull operator to use its interface for withdrawals or swaps becomes part of the fraud infrastructure – and faces potential compliance exposure under FATF guidelines and MiCA regulations.

ChainAware’s DApp protection layer screens connecting wallets at the moment of wallet connection – before any transaction is submitted. Wallets associated with known rug pull contracts, flagged behavioral patterns, or AML-positive addresses are identified at the connection stage and can be blocked, flagged for review, or routed to restricted functionality automatically. The full architecture of this pre-connection screening is covered in our guide to Web3 fraud detection for DApps in 2026.

For DApps operating under MiCA requirements, ChainAware’s compliance layer provides 70-75% MiCA coverage at approximately 1% of the cost of traditional compliance solutions from Chainalysis or Elliptic. The complete comparison of DeFi compliance tools and costs is available in our DeFi compliance tools comparison for 2026.

The AML Layer: Complementary, Not Overlapping

AML screening and rug pull detection address different risk vectors. AML screening asks: is this wallet associated with known illicit activity – sanctions lists, mixer usage, dark web transactions? Rug pull detection asks: is this contract creator likely to extract liquidity before investors can exit?

These questions are complementary. A wallet can be AML-clean but behaviorally likely to rug – a fresh wallet with no prior illicit associations but a clear deployment pattern matching rug pull operators. Conversely, a wallet can trigger AML flags without being involved in rug pulls – a legitimate DeFi user who passed through a mixer for privacy reasons, for instance.

Running both checks provides the most comprehensive protection. For the complete technical architecture of AML and KYT compliance for DeFi, our guide to blockchain compliance, KYT, and AML for DeFi in 2026 covers the regulatory obligations and implementation options for protocols of all sizes. Additionally, for protocols evaluating predictive AI versus traditional rule-based approaches for compliance, our analysis of predictive AI for crypto KYC, AML, and transaction monitoring provides a direct comparison.

PancakeSwap V2 Specifically: Why BNB Chain Is Ground Zero for Rug Pulls

PancakeSwap V2 on BNB Chain represents an ideal operational environment for rug pull schemes. Understanding why helps investors calibrate their risk exposure appropriately when interacting with BNB Chain DeFi specifically.

BNB Chain gas fees are among the lowest of any major EVM-compatible chain. Deploying a token contract, creating a PancakeSwap V2 pool, and seeding initial liquidity can cost less than $5 in total gas fees. This near-zero barrier to entry means that fraud operators can deploy hundreds of rug pull setups simultaneously at negligible marginal cost. A single successful extraction – even of a few thousand dollars – covers the cost of dozens of attempted schemes.

PancakeSwap V2 also benefits from extremely high organic retail traffic. Hundreds of thousands of retail investors use PancakeSwap V2 daily, creating a large pool of potential victims for each fraudulent token. The PancakeSwap interface itself presents tokens without any fraud risk warnings – it is a neutral trading interface, not a security layer. The responsibility for fraud detection falls entirely on the investor.

Token aggregators like DexTools and DexScreener surface newly created pools with real-time price charts, trading volume, and holder counts – all of which can be manipulated through bot trading and wash volume. A fresh pool with manufactured trading activity looks identical to a legitimate new project on these interfaces. Retail investors using aggregators as their primary research tool are working with data that fraud operators have specifically optimized to deceive.

CertiK’s 2025 security data placed BNB Chain consistently among the top chains for exit scam and rug pull activity, a position it has held across multiple annual reporting periods. The combination of low costs, high traffic, and minimal native protection makes BNB Chain the most active venue for the rug pull industry globally.

This is not a criticism of BNB Chain or PancakeSwap as infrastructure. Both are legitimate, high-quality platforms. The fraud problem is an emergent consequence of their success – high retail traffic and low costs that serve legitimate users equally serve fraudulent ones.

What the Rug Pull Industry Looks Like at Scale

$569M across 20 weeks. 103,695 individual events. These numbers only make sense at the scale of an organized industry, not at the scale of individual bad actors. Consider the operational requirements to produce this output.

At an average of 5,185 rug pull events per week, operators must be deploying thousands of token contracts weekly. Each contract requires wallet funding, deployment, pool creation, liquidity seeding, marketing (to attract retail buyers), and eventually liquidity removal. The automation required to manage this volume is sophisticated – these are not manual operations but scripted, bot-driven pipelines that handle the entire lifecycle with minimal human intervention.

The diversity of fraud values also tells a story. The same week that produced the maximum fraud value ($53.4M in W4) contained individual rug pulls ranging from a few hundred dollars to millions. This range reflects an ecosystem with multiple tiers: small-scale operators running low-value schemes alongside professional operations running high-value targeted campaigns. The industry has a hierarchy, with the most sophisticated operators at the top extracting the most per event while the smallest operators run volume plays at low margins.

Understanding rug pulls as an industry – not as isolated frauds – changes how we think about protection. Blacklists of known bad actors are largely ineffective against industrial-scale operations that create fresh wallets continuously. Reactive forensics – identifying fraud after it happens – provide no protection to the retail investors who lost money. Only predictive, behavioral approaches that identify fraud operators before they extract value offer meaningful protection at this scale.

This is the core thesis behind ChainAware’s approach. For context on how predictive AI compares to forensic analytics in practical terms, and why the shift from reactive to predictive is the fundamental security challenge facing Web3 in 2026, our comparison of forensic versus AI-powered blockchain analysis provides the complete framework.

The Role of Token Holder Quality in Identifying Rug Pull Risk

One additional signal that Rug Pull Detector V3 incorporates – particularly for tokens that have already launched and have a holder base – is token holder quality analysis. ChainAware’s Token Rank system assigns every token a rank based on the median Wallet Rank of its holders. Legitimate tokens with real communities tend to attract wallets with diverse, experienced behavioral profiles. Rug pull setups, which attract retail speculators and bots, produce distinctive holder quality signatures.

A token where the holder base consists predominantly of fresh wallets with no prior DeFi history, created in the days immediately preceding the token launch, is a warning sign that the holder base was manufactured rather than organically acquired. Bot-driven trading activity similarly produces holder clusters with synchronized creation dates and homogeneous transaction histories – patterns that stand out clearly in a holder quality analysis.

For investors conducting due diligence on tokens with existing holder bases, ChainAware’s Token Rank guide covers the complete methodology for assessing holder quality as part of investment due diligence. Combined with Rug Pull Detector V3’s contract and creator analysis, holder quality analysis provides a third independent signal layer – and convergent warnings across all three layers are among the strongest indicators of fraudulent intent available in the DeFi ecosystem today. This deployer- and holder-level scoring is the same methodology behind ChainAware’s specialised ready-made agents, which combine token risk with holder and deployer screening in a single call.

P2P Transactions: The Rug Pull Risk Outside DEX Pools

The $569M we measured operates entirely within the DEX pool context – tokens and liquidity pools on PancakeSwap V2. However, rug pull risk also exists in peer-to-peer payment contexts that do not involve DEX pools at all.

Approximately 50% of on-chain transaction volume consists of direct peer-to-peer transfers – one wallet sending assets directly to another, with no DApp interface in the flow. This volume includes legitimate payments, OTC trades, escrow arrangements, and investment contributions to projects that have not yet launched on a DEX. It also includes fraud: investors sending funds to project team wallets that subsequently disappear, or OTC trades where the receiving party does not fulfill their side of the arrangement.

For the 50% of transactions that happen wallet-to-wallet, the protection question is not “is this pool safe?” but “is this wallet safe?” ChainAware’s free Wallet Auditor addresses exactly this use case – providing a complete behavioral profile and fraud risk assessment for any wallet address before you send irreversible on-chain funds to it. The Wallet Auditor complete guide walks through every parameter the audit covers and how to interpret the results for P2P transaction due diligence.

What Comes Next: ChainAware’s Rug Pull Research Roadmap

This dataset – 103,695 rug pulls and $569M in confirmed extraction – is the foundation of an ongoing research program. Publishing this data is the beginning, not the conclusion.

The next phases of ChainAware’s rug pull research and product development address several open questions that this initial dataset raises but does not answer.

Expanding to Complex Rug Pull Detection

As noted throughout this article, the $569M figure covers only basic liquidity extraction. Adding LP token transfer rug pulls, unlocked token sell-offs, and associated wallet extraction to the detection methodology will require additional algorithmic development – but the training data now exists in our confirmed dataset. V3.1 and subsequent versions will incrementally expand coverage to these more complex patterns.

Multi-Chain Expansion

This dataset covers PancakeSwap V2 on BNB Chain. ChainAware already operates across 8 blockchains – ETH, BNB, BASE, POLYGON, SOL, TON, TRON, and HAQQ. Expanding the rug pull dataset to include Ethereum DEXes (Uniswap V2/V3), Solana meme token launchpads, and BASE chain activity will produce a comprehensive multi-chain picture of rug pull extraction that does not currently exist anywhere in the industry.

Solana’s pump.fun – where an estimated 99% of tokens launched end in loss events for buyers – is a particularly important next target for this analysis framework. The structural characteristics of pump.fun token launches share significant overlap with PancakeSwap V2 rug pulls, but the technical analysis requires adaptation to Solana’s account-based architecture and SPL token standard.

Continuing Algorithm Improvement

The 90.1% accuracy of V3 is a milestone, not a ceiling. The algorithm team continues iterating on both the behavioral and smart contract analysis pipelines. Each confirmed rug pull event from our ongoing monitoring adds to the training dataset, and each fraud industry adaptation that our current model misses provides labeled false negative examples for the next training cycle.

The target is 95%+ prediction accuracy. Reaching that threshold requires solving the core challenge of sophisticated operator evasion – fraud operators who invest significantly in maintaining clean behavioral profiles and deploying contracts that pass automated inspection. ChainAware’s research team is actively working on graph-based analysis of funding networks and deployment clustering to address this remaining gap.

Rug Pull Red Flags: What to Look for Before Investing in Any New Token

While automated tools like V3 do the heavy lifting, investors who understand the underlying red flags make better decisions – both by interpreting V3’s output intelligently and by performing manual checks that complement automated analysis. The following red flags are drawn directly from our analysis of the 103,695 rug pull events in the dataset and from ChainAware’s behavioral research across 20M+ wallet profiles.

Red Flag 1: Unrenounced Contract Ownership

Contract ownership allows the deployer to call privileged functions – modifying transfer fees, adding transfer restrictions, minting additional supply, and pausing trading. A legitimate project with long-term intentions almost always renounces ownership or transfers it to a time-locked multisig. A contract where the deployer retains full ownership indefinitely is structurally set up for the owner to modify conditions after retail investors have bought in.

Checking ownership renouncement takes 30 seconds on BscScan or Etherscan – search the contract address, look for the “Contract” tab, and check whether owner() returns the zero address (renounced) or an active wallet address (retained). This single check eliminates a significant fraction of the most blatant rug pull setups.

Red Flag 2: Unlocked Liquidity Pool Tokens

When a creator seeds liquidity into a PancakeSwap V2 pool, they receive LP (Liquidity Provider) tokens representing their share of the pool. Legitimate projects lock these LP tokens in a time-lock contract – preventing liquidity removal for a defined period (commonly 6-24 months). Unlocked LP tokens held in the deployer’s wallet can be redeemed for the underlying liquidity at any moment, making rug pull execution a one-transaction event.

LP lock verification can be performed manually on platforms like Mudra or Team Finance, which maintain records of locked LP tokens on BNB Chain. The absence of a lock is not automatically fraudulent – some legitimate projects rely on other mechanisms – but it is a significant red flag that requires additional due diligence before committing capital. V3’s smart contract analysis flags unlocked LP positions automatically.

Red Flag 3: Fresh Deployer Wallet with No History

Professional rug pull operators often maintain separate deployer wallets for each scheme – fresh addresses with no prior on-chain history. This is a deliberate evasion tactic: behavioral analysis tools that look at deployer history return a clean result, because there is no history to analyze. V3 handles this by examining the funding source of the deployer wallet (where the ETH or BNB originated from), not just the deployer wallet itself. However, a fresh deployer wallet with no history should still independently raise an investor’s suspicion level, particularly when combined with other red flags.

Red Flag 4: Suspiciously High Maximum Transaction or Wallet Limits

Many rug pull contracts include maximum transaction size or maximum wallet balance limits that restrict how much any individual buyer can purchase. These limits are presented as anti-whale measures – preventing any single buyer from acquiring too large a share. In practice, they often serve a different purpose: ensuring that no single buyer has enough exposure to justify the gas cost of legal action, and ensuring that the token requires extended time to accumulate retail capital before the rug can be executed. Combined with gradual price increases driven by bot activity, these limits produce a slow accumulation of widely distributed retail capital that is then extracted in a single pull event.

Red Flag 5: Concentrated Token Allocation

Token distribution matters enormously. If the top 10 wallets hold more than 50% of the total supply, and those wallets are connected to the deployer or hold unlocked tokens, they represent a latent supply overhang that can be sold into any retail buying pressure. ChainAware’s Token Rank analysis reveals the quality and concentration of holder distribution for any token with an existing holder base – a concentrated, low-quality holder base is one of the strongest predictors of eventual rug pull or dump events. The complete methodology is explained in the Token Rank guide.

Red Flag 6: Anonymous Team With No Verifiable Presence

Full anonymity is the norm in crypto, not the exception – many legitimate projects have anonymous teams. However, the combination of full anonymity with the absence of any verifiable prior project history, a project launched in the previous week with no audit, and an aggressive social media presence focused entirely on price expectations is a signature pattern of rug pull operations. Legitimate anonymous projects typically have verifiable GitHub commit histories, prior community involvement in legitimate projects, and at minimum a third-party security audit. The absence of all three simultaneously is a meaningful red flag.

Red Flag 7: Extremely High Early Price Performance

A token that gains 500%-2,000% in its first 24-48 hours of trading is an appealing investment narrative. It is also the characteristic signature of a pump phase that precedes a rug pull. Organic price discovery – driven by genuine retail demand for a real product – produces growth curves that are steep but not parabolic. Parabolic early gains are typically produced by bot-driven wash trading that manufactures artificial volume and price momentum to attract FOMO-driven retail buyers. The parabolic chart is the marketing material for the rug. When the price is high enough to make liquidity removal maximally profitable, the rug executes.

How Rug Pull Operators Evade Detection: An Adversarial Analysis

Understanding how sophisticated fraud operators attempt to evade detection is as important as understanding the detection methods themselves. The 9.9% of rug pull events that V3 does not catch are concentrated in operations that specifically invest in evasion. Publishing this analysis serves retail investors by explaining what the highest-risk events look like, and serves the research community by documenting the evasion tactics that future algorithm versions must address.

Wallet Aging and History Manufacturing

The most sophisticated operators maintain aged deployer wallets – addresses that have months or years of legitimate-looking transaction history before being used for fraud. These wallets interact with legitimate DeFi protocols, make small token purchases across diverse projects, and accumulate behavioral signals that behavioral analysis models associate with legitimate users. When such a wallet eventually deploys a fraudulent contract, the behavioral layer returns a low-risk signal. Only the smart contract analysis layer can catch fraud from these operators – and only if the contract itself contains detectable risk patterns.

Countering this tactic requires tracking the clustering of aged wallets that share funding sources or that appear in coordinated deployment patterns – even when each individual wallet’s history appears clean. This graph-based analysis is part of ChainAware’s ongoing algorithm development roadmap.

Contract Code Obfuscation

Professional fraud operators increasingly deploy contracts with obfuscated code – transfer restrictions hidden within complex modifier chains, mint functions embedded in proxy contract architectures, and ownership retention disguised as multi-sig governance mechanisms that are controlled entirely by the deployer. These obfuscation patterns specifically target the limitations of automated smart contract analysis tools.

V3’s bytecode inspection – which operates on compiled contract code rather than requiring readable source code – provides partial protection against source code obfuscation. However, novel obfuscation patterns that have not yet appeared in the training dataset represent genuine blind spots. Regular model updates and the continuous addition of newly confirmed rug pull events to the training dataset are the primary mechanism for closing these gaps as they are discovered.

Delayed Execution

Some operators run long-duration schemes – maintaining active development activity, social media presence, and liquidity for weeks or months before executing the rug pull. These delayed execution schemes are specifically designed to outlast the attention spans of initial investors who may have checked V3 at launch and seen a clean result. By the time the rug executes, many initial investors have already forgotten the risk assessment they saw weeks ago and have added to their position based on the apparent ongoing legitimacy of the project.

Protecting against delayed execution requires ongoing monitoring rather than a single point-in-time check. ChainAware’s Transaction Monitoring Agent, which continuously rescreens connecting wallets on every DApp visit, provides this ongoing monitoring layer for DApps. For individual retail investors, re-running V3 checks periodically on held positions – particularly when planning to add to an existing position – is the equivalent individual protection behavior.

Legitimate-Looking Audits

A concerning development in the rug pull industry is the emergence of fraudulent security audits – certificates from obscure or non-existent “audit firms” that create the appearance of third-party verification without the substance. A token that displays an “audited” badge from an unrecognizable firm – or from a legitimate-sounding name that does not correspond to any established security company – provides no real protection against rug pull risk.

Legitimate audits from established firms like CertiK, Trail of Bits, Consensys Diligence, or OpenZeppelin are verifiable on the auditing firm’s own website and are associated with published audit reports. An audit badge that cannot be verified against the auditing firm’s own published report list should be treated as fabricated. V3’s smart contract analysis provides an independent assessment that does not rely on audit claims – it examines the code directly, regardless of what audit certificate the project displays.

The Market Infrastructure Gap: Why DeFi Needs Better Fraud Data

The $569M figure in this report exists not just as a fraud statistic but as a market infrastructure data point. DeFi cannot mature into a mainstream financial system while retail investors lose hundreds of millions of dollars per quarter to a fraud mechanism that could be largely prevented with better tooling and better data.

Traditional financial markets have infrastructure specifically designed to protect retail investors from comparable fraud: prospectus requirements, securities registration, market maker oversight, exchange listing standards, and regulatory enforcement. These mechanisms are not perfect – traditional markets have their own fraud problems – but they represent decades of accumulated institutional knowledge about how to structure market access to minimize retail harm.

DeFi has none of this infrastructure by design – permissionless access is a core feature, not a bug. The trade-off is that permissionless access enables both the extraordinary innovation of the DeFi ecosystem and the extraordinary fraud of the rug pull industry. Building protective infrastructure that preserves permissionlessness while reducing retail harm requires data-driven tools that operate at the protocol layer – not regulatory gatekeeping that would undermine DeFi’s fundamental architecture.

ChainAware’s approach – behavioral AI that flags risk without blocking access, tools that inform investor decisions without requiring permission from any central authority – represents one model for how this protective infrastructure can be built. Publishing the $569M dataset is part of the broader argument that this infrastructure is urgently needed and that the data to build it exists.

For DApp teams specifically, building fraud resistance into their platform architecture is becoming a competitive differentiator. Users who have been burned by rug pulls – and statistically, a significant fraction of active DeFi users have been – actively seek platforms with visible security measures. A DApp that visibly screens connecting wallets, displays behavioral security ratings for tokens listed on its interface, and provides transparent fraud risk data builds trust with the retail user base that DeFi needs to grow beyond its current audience.

The DeFi onboarding problem – why 90% of connected wallets never transact – is partly a product problem, partly a UX problem, and significantly a trust problem. Users who don’t trust that their capital is safe don’t commit capital. Solving fraud at the infrastructure level directly addresses the trust component of the conversion problem. For the complete analysis of why the 90% non-transacting wallet problem is DeFi’s most critical growth challenge, and how behavioral intelligence addresses it, see our comprehensive guide on DeFi onboarding and why connected wallets don’t transact.

Building that trust requires two things simultaneously: making fraud harder to execute and making fraud more visible when it does execute. V3 contributes to making fraud harder to execute – at 90.1% prediction accuracy, it prevents a significant fraction of rug pull investments before they happen. Publishing this dataset contributes to making fraud more visible – giving the industry, regulators, and researchers the empirical foundation to understand and respond to the true scale of the problem.

Together, these two contributions represent ChainAware’s core thesis: that Web3 grows faster and serves users better when behavioral intelligence is embedded into the infrastructure of every DApp interaction. The $569M figure is not just a warning – it is the business case for why predictive fraud detection is one of the most important infrastructure investments any Web3 platform can make in 2026. For the complete framework of how behavioral intelligence powers Web3 growth beyond security, the Web3 agentic economy guide covers the full stack from fraud prevention through personalized growth. The full 19-category forensic methodology behind the Fraud Detector is documented on the Fraud Detector learn page.

Frequently Asked Questions

What exactly is a rug pull?

In the context of this dataset, a rug pull is defined as a liquidity event where the contract creator of a token pool removes more liquidity than they originally added. A creator deposits funds (Mint event), retail investors buy the token and add value to the pool, and then the creator withdraws all liquidity (Burn event). The difference between what they removed and what they added – when removal exceeds addition – is the rug pull value. The token price drops to zero instantly upon liquidity removal, leaving holders with worthless assets. For a full comparison of rug pulls versus pump-and-dump schemes, see our dedicated guide on rug pull vs pump and dump.

How does ChainAware Rug Pull Detector V3 work?

V3 runs two parallel analysis pipelines. The first examines the behavioral history of the contract creator’s wallet – their deployment history, funding sources, prior rug pull associations, and wallet age. The second inspects the smart contract code itself for dangerous patterns: hidden transfer restrictions, uncapped mint functions, unrenounced ownership, and unlocked LP tokens. Both pipelines produce independent risk scores that are combined through an ensemble model trained on 103,695+ confirmed rug pull events. The combined model achieves 90.1% prediction accuracy on the held-out test set.

Is the Rug Pull Detector free?

Yes. The web interface at chainaware.ai/rugpull is completely free for retail investors, with no account creation required. Businesses that need API access for automated, high-volume token screening can subscribe to the API at chainaware.ai/subscribe. AI agents can access the tool through the X402 micropayment protocol, paying per analysis without requiring a subscription.

What is the difference between V2 and V3?

V2 relied exclusively on behavioral analysis of contract creator wallets, achieving approximately 68% prediction accuracy. V3 adds a full smart contract analysis layer – inspecting contract code for dangerous patterns regardless of the deployer wallet’s history. This combination closes the gap that sophisticated fraud operators exploited in V2 by maintaining clean deployer wallet histories. The combined V3 model achieves 90.1% prediction accuracy – a 32.5% relative improvement over V2.

Does V3 work on tokens that have not yet launched?

Yes. V3 can analyze any deployed smart contract – including contracts that have been deployed but have not yet attracted liquidity. Smart contract analysis runs regardless of whether there is an active trading pool. Creator behavioral analysis also runs at any point after contract deployment. For pre-launch tokens, smart contract analysis is particularly valuable because it does not require any trading history to return a risk assessment.

Can a rug pull still happen even if V3 gives a low risk score?

Yes. V3 achieves 90.1% accuracy, meaning approximately 9.9% of rug pull events will not be flagged. These false negatives are concentrated in the most sophisticated fraud operations – those that invest in maintaining clean deployer profiles and deploy contracts that pass automated inspection. No automated tool can guarantee 100% detection. V3 should be used as a risk filter, not as a guarantee, and combined with human judgment on tokenomics, team identity, and project fundamentals for high-value investment decisions.

Why does $569M in losses generate almost no media coverage?

Rug pull losses are distributed across hundreds of thousands of small transactions affecting individual retail investors, rather than concentrated in a single event affecting an institution. Distributed losses are structurally invisible to media coverage, which requires a single dramatic event with a named victim and a named perpetrator. The victims of rug pulls are anonymous retail investors with no media presence. The perpetrators are anonymous wallet operators with no attributable identity. The structural invisibility of rug pull losses is one of the primary reasons this form of fraud operates at scale without triggering the industry-wide response that dramatic exchange hacks generate.

What chains does ChainAware support for rug pull detection?

ChainAware currently supports 8 blockchains across its full product suite: Ethereum (ETH), BNB Smart Chain (BNB), Base (BASE), Polygon (POLYGON), Solana (SOL), TON, TRON, and HAQQ. Rug Pull Detector V3 is optimized for BNB Chain and Ethereum in its current version, with Solana and Base coverage expanding in upcoming releases. For the complete multi-chain capability overview, ChainAware’s complete product guide covers which tools are available on which chains.

How can AI agents use Rug Pull Detector V3?

AI agents access Rug Pull Detector V3 through ChainAware’s X402 micropayment protocol – paying per analysis in real time without requiring pre-approved API keys. Agents can also access the full suite of ChainAware tools through the Model Context Protocol (MCP) server at prediction.mcp.chainaware.ai/sse. The complete integration guide for AI agents is available at our Prediction MCP guide, and the GitHub repository at github.com/ChainAware/behavioral-prediction-mcp provides open-source agent definitions that any AI framework can deploy directly.

The post $569M+ in Rug Pulls on PancakeSwap V2 in 20 Weeks – Rug Pull Detector V3 Launched With 90.1% Accuracy first appeared on ChainAware.ai.