| 1 | | A- | A- | L1 | $37.7B | Staking centralization — top entities including Lido (~25%) and major CEXs collectively control >60% of staked ETH; below the 33% single-entity censorship threshold but the concentration creates cartelization and correlated-failure risk | ▼ 3 |
| 2 | | B | B | Liquid Staking | $17.8B | ~23% of all staked ETH controlled by one protocol creates Ethereum-level systemic centralization risk; market share has declined from a 2023 peak of ~32%, reducing proximity to the 33% consensus interference threshold | → 0 |
| 3 | | C | B+ | Lending | $13.9B | Accepted stolen rsETH as e-mode collateral on April 18, 2026 after Kelp DAO's LayerZero bridge was exploited for $292M; attacker borrowed WETH against now-worthless collateral, leaving Aave V3 with $177-200M in bad debt. WETH pool hit 100% utilization, $6.2B in withdrawals, AAVE -17.7%. Recovery plan: a DeFi United coalition (Consensys, Lido, EtherFi, others) pledged $300M+ and published a technical proposal on April 28 to restore full rsETH backing without socializing losses; governance votes across Ethereum and Arbitrum deployments are pending as of April 30. Umbrella holds ~23,500 WETH (~$54M), covering ~60% of L1 Core losses; outstanding deficit contingent on vote outcome. | ▼ 2 |
| 4 | | B | C+ | Liquid Staking | $9.4B | DVT splits validator keys across 4+ operators via Shamir Secret Sharing — a compromised threshold (3-of-4) of operators could forge attestations or double-sign, risking slashing of the 7.4M+ ETH secured by SSV (~118,000 validators). | ▲ 2 |
| 5 | | C | C | Lending | $7.8B | Oracle-agnostic permissionless markets: any oracle can be used in any market, including low-quality or malicious ones; the October 2024 PAXG exploit ($230K) confirmed oracle misconfiguration is the primary attack vector | — |
| 6 | | B- | C | Lending | $7.6B | P2P matching engine adds complexity: if matching fails, fallback to pool rates may surprise users | ▼ 2 |
| 7 | | C | B | Bridge | $7.4B | Custodial 2-of-3 key concentration in Justin Sun entities: BiT Global (Justin Sun-affiliated, HK) now holds two of the three keys in WBTC's custody vault arrangement — the user key (Hong Kong) and backup key (Singapore) — while BitGo retains one key and provides infrastructure. A BiT Global insolvency, regulatory seizure, or governance dispute could impair the 1:1 BTC backing on ~$7.4B of circulating WBTC. BitGo's OCC bank charter and Fortune 500 status (NYSE: BTGO) provide a strong institutional backstop on the BitGo side. | ▲ 1 |
| 8 | | B- | D- | Liquid Staking | $7.1B | Centralized custody: all staked ETH is managed by Binance validators, creating a single-entity dependency for ~$7.4B in assets | ▲ 2 |
| 9 | | C- | C+ | Bridge | $6.7B | Infrastructure compromise confirmed: the April 2026 KelpDAO exploit ($292M) demonstrated that social engineering of LayerZero developers combined with DDoS-forced RPC failover can bypass DVN verification for single-verifier (1-of-1) applications. LayerZero admitted fault on May 9, 2026; $2B+ in TVL migrated to competing solutions. | ▲ 9 |
| 10 | | B- | B- | L1 | $6.2B | Network reliability — history of extended outages requiring validator coordination to restart | ▼ 2 |
| 11 | | B | B | CDP | $6.2B | Oracle-dependent liquidation system: Maker relies on a custom oracle module (Medianizer/OSM with 1-hour delay) feeding ETH and other collateral prices. During Black Thursday (March 2020), oracle lag combined with network congestion led to $8.3M in zero-bid liquidation auctions. The system has since been rebuilt with Liquidations 2.0 (Dutch auction format) and Chainlink integration, substantially mitigating but not eliminating oracle-related liquidation risk. | ▼ 5 |
| 12 | | B- | A- | CDP | $5.7B | USDS freeze function introduces censorship risk that undermines decentralization, splitting the community between DAI purists and USDS adopters | → 0 |
| 13 | | C- | B+ | CDP | $5.7B | Extremely thin capital buffer: S&P rated the protocol B- citing a 0.4% capital ratio — any significant bad-debt event could threaten solvency | — |
| 14 | | C- | B | Derivatives | $5.2B | CME Group and ICE formally lobbied CFTC (May 2026) to mandate Hyperliquid register as a derivatives exchange with KYC and trade surveillance; the platform was also used as a laundering passthrough in the May 2026 THORChain exploit, reinforcing the sanctions evasion narrative regulators have advanced | ▼ 1 |
| 15 | | C+ | C+ | L1 | $5.2B | Centralization — only 21 cabinet validators produce blocks (45 total including candidates), all effectively controlled by Binance ecosystem | ▼ 2 |
| 16 | | B- | C- | Restaking | $5.1B | EigenLayer introduced restaking as a novel mechanism category where staked ETH simultaneously secures multiple Actively Validated Services (AVSs), creating correlated slashing risk — an operator slashed on one AVS could trigger cascading unstaking across other AVSs they secure, though the April 2025 slashing upgrade introduced unique allocated stake per AVS to contain blast radius. | → 0 |
| 17 | | B | C+ | L1 | $5.1B | Rainberry Inc. (Tron-associated) settled SEC charges for $10M in March 2026 with all claims dismissed with prejudice, removing the primary regulatory overhang. Justin Sun's dual federal litigation — suing WLFI for fraud (alleging governance manipulation over his $107M position) while being counter-sued by WLFI for defamation and market manipulation — both cases active in U.S. federal court as of July 2026, introducing fresh reputational and legal risk tied to the key person. | ▲ 1 |
| 18 | | C+ | C+ | Restaking | $5.1B | Protocol generates minimal organic revenue — the $541K/month ($6.5M/year) in real fees is dwarfed by ~$12M/year in EIGEN emissions; ELIP-012 (approved March 2026) routes real fees to buyback but net dilution continues as emissions exceed revenue | ▼ 1 |
| 19 | | C | B | Stablecoin | $3.8B | Reserve fund ($62M) covers ~1.4% of $4.4B USDe supply — depletes in ~52 days under the protocol's own V1 stress test at -10% annualized funding (coverage ratio thinned further as supply recovered from the $3.8B trough to a May 2026 high near $5.4B, only to contract again as funding rates compressed) | ▼ 2 |
| 20 | | B | B | Lending | $3.7B | Deep dependency on Sky (MakerDAO) ecosystem: protocol solvency is backstopped by Sky's $6.5B reserve, creating single-entity systemic risk | ▼ 1 |
| 21 | | B- | C- | Lending | $3.7B | Sky ecosystem concentration: SparkLend is deeply coupled to MakerDAO/Sky governance and the USDS/DAI stablecoin. SparkLend's liquidity backbone (the Spark Liquidity Layer) is Sky-controlled — a governance crisis or DAI instability would directly impair SparkLend's lending capacity. | ▼ 2 |
| 22 | | B- | C | RWA | $3.5B | Multi-chain bridge risk: BUIDL deploys across Ethereum, Solana, Polygon, BNB Chain, and Avalanche via Wormhole; a bridge exploit could mint unbacked tokens or freeze legitimate holders' assets across chains | ▼ 3 |
| 23 | | C | B | Bridge | $3.4B | A bug in the OFT contracts or collusion among 2-of-3 Decentralized Verifier Networks (DVNs) could enable unbacked USDT0 minting, creating synthetic supply not backed by locked USDT in the Ethereum lockbox. Nine independent audits and a $6M Immunefi bug bounty reduce but do not eliminate this risk. | — |
| 24 | | C | B- | Liquid Staking | $3.3B | EigenLayer restaking slashing (conditional risk): eETH validators are set up for EigenLayer restaking, but ether.fi operators have not yet opted into slashing modules per Chaos Labs' governance analysis. When/if operators opt in, all eETH holders would share proportional losses from any AVS slashing event — the architecture supports socialized loss even though the risk is not yet active. | ▼ 3 |
| 25 | | C- | B- | Restaking | $3.3B | EigenLayer restaking with socialized slashing: all eETH holders share proportional losses if an AVS is slashed. EigenLayer's live slashing system (since April 2025) makes this an active risk — a major AVS incident could reduce eETH's value for all holders simultaneously. | ▲ 2 |
| 26 | | B | C- | L2 | $3.2B | The Security Council (9-of-12 multisig) can perform emergency upgrades to all Arbitrum contracts without any timelock delay, creating a centralization risk where a compromised or coerced council could alter the rollup's behavior instantly. The DAO has published the council member identities and an election process to mitigate this. In April 2026, the Security Council exercised this freeze power to lock $71M in ETH tied to the Kelp DAO exploit, demonstrating real-world use of this emergency authority. | ▼ 2 |
| 27 | | B- | D | RWA | $3.0B | USYC is a permissioned, KYC-gated token representing the Hashnote International Short Duration Yield Fund. Regulatory changes to tokenized securities could force redemption freezes or operational changes, with $1.7B in assets at risk. | → 0 |
| 28 | | C+ | B+ | RWA | $3.0B | USYC is now a Circle product (acquired Jan 2025, Circle NYSE:CRCL since June 2025), improving regulatory standing, but ~65% of USYC TVL (~$1.84B) is concentrated on BNB Chain as Binance institutional collateral via Ceffu custody — a new single-counterparty concentration risk. | ▼ 1 |
| 29 | | B | B- | DeFi | $2.9B | Curator misallocation risk — Steakhouse allocates $2.9B+ across vaults, and a single bad market selection could cascade across all positions. The March 2026 Resolv Labs hack ($23M via compromised private key) occurred while Steakhouse served as Resolv's risk manager; Steakhouse had correctly flagged the attack vector 5 days prior and held zero direct vault exposure — but the incident confirms that curator oversight cannot substitute for client protocol key management hygiene | ▼ 2 |
| 30 | | C+ | A- | RWA | $2.9B | Tether corporate contagion risk: despite separate legal structure, XAUt's association with Tether (USDT issuer) creates reputational and regulatory risk if parent company faces enforcement actions or banking failures | ▼ 3 |
| 31 | | B- | C+ | L2 | $2.8B | Coinbase is sole sequencer with no permissionless fallback, creating a corporate single point of failure for $4.1B in TVL — though Stage 1 decentralization (Jan 2026) now allows users to exit without sequencer cooperation. | → 0 |
| 32 | | C | B | RWA | $2.6B | Single custodian concentration: all $2.64B in backing assets (US Treasuries, BlackRock iShares ETF, bank deposits) are held by Morgan Stanley Smith Barney LLC. A custodian failure, freeze, or regulatory action against the custodian would prevent USDY redemptions, with no backup custodian arrangement disclosed. | — |
| 33 | | C+ | B+ | RWA | $2.6B | Counterparty risk on underlying custodians and fund managers — if short-term Treasury backing fails, USDY depegs | → 0 |
| 34 | | C- | B- | Restaking | $2.6B | BLS vote extension vulnerability allows validators to bypass consensus by omitting block hash fields, undermining the security model at its core. | ▲ 1 |
| 35 | | C- | C | Restaking | $2.6B | Self-custodial BTC staking via EOTS is a novel cryptographic primitive with no battle-tested precedent — any flaw in slashing/extraction logic could irreversibly forfeit staked BTC | ▼ 1 |
| 36 | | C- | C | Yield | $2.5B | New protocol with large TVL and limited track record: Grove Finance launched in June 2025 and has rapidly accumulated $2.7B TVL. No audits are publicly documented in DeFiLlama's data. A protocol of this size with less than one year of track record and unverified audit status represents meaningful smart contract risk. | — |
| 37 | | B | C- | RWA | $2.2B | Spiko tokenizes money market funds backed by US and EU Treasury bills — while the underlying assets are low-risk, the tokenization layer introduces smart contract, custody, and regulatory surface area that traditional T-bill investors don't face. | ▼ 4 |
| 38 | | B- | D+ | DeFi | $2.2B | Kraken DeFi Earn concentration: Kraken's integration as the primary TVL driver means a platform withdrawal or regulatory action affecting Kraken could force rapid liquidation of $500M+ in DeFi positions at distressed prices | → 0 |
| 39 | | C- | C | Lending | $2.2B | Core Foundation obtained a Cayman Islands court injunction (March 2026) blocking Maple from launching syrupBTC, alleging misuse of confidential information from their joint lstBTC development — this blocks a $150M+ institutional asset product and introduces legal/operational overhang. | ▲ 1 |
| 40 | | B+ | A- | RWA | $1.8B | BUSD wind-down precedent: Paxos was forced by NYDFS to cease BUSD operations in 2023, demonstrating that even federally chartered products can be shut down by regulators — a risk class that applies to PAXG. | → 0 |
| 41 | | B | B- | DEX | $1.7B | Dominant BSC DEX position creates systemic concentration risk; BSC chain-level issues directly impact ~$1.7B TVL | → 0 |
| 42 | | B- | C- | Yield | $1.7B | Capital deployed across multiple chains and DeFi protocols means a failure in ANY recipient protocol cascades losses back through the entire Spark/Sky ecosystem | ▼ 2 |
| 43 | | B- | C+ | RWA | $1.6B | Real-world asset counterparty and default risk is inherently opaque on-chain; 2023 default event exposed originator vetting weaknesses | ▲ 1 |
| 44 | | C+ | C- | RWA | $1.6B | February 2026 data breach compromised 967,000 user records via Okta SSO social engineering — no on-chain impact but elevates regulatory scrutiny and customer phishing exposure | ▼ 1 |
| 45 | | C | C- | DEX | $1.6B | Order matching is centralized (off-chain, operated by Figure Markets) despite on-chain settlement: if Figure's matching engine goes offline or is compromised, users cannot execute trades even though their assets remain secure in MPC wallets — creating a critical dependency on Figure Technologies as a corporation. | — |
| 46 | | B | B+ | DEX | $1.5B | Concentrated liquidity amplifies impermanent loss when prices move out of LP-set ranges | ▲ 4 |
| 47 | | C+ | C- | Bridge | $1.5B | February 2022 exploit allowed minting 120,000 wETH ($320M) without collateral via signature verification bug; Jump Crypto backstopped losses | ▼ 5 |
| 48 | | D+ | D+ | Restaking | $1.5B | LBTC is a bridge-dependent wrapped BTC derivative — the KelpDAO April 2026 $292M LayerZero exploit directly templates the attack vector for LBTC (multi-chain OFT-style distribution, bridge config risk) | — |
| 49 | | C+ | C+ | DeFi | $1.4B | Gauntlet's simulation-based risk models curate $2B+ in vault AUM and inform parameters for protocols with $35B+ in monitored assets — models calibrated on historical data may fail catastrophically during tail events outside observed volatility ranges | ▲ 1 |
| 50 | | B | C+ | Yield | $1.3B | Vault curator model introduces principal-agent risk — curators allocate capital across DeFi strategies on behalf of depositors | ▼ 2 |
| 51 | | B | B | DEX | $1.3B | Vyper compiler vulnerability (July 2023 exploit) eroded trust; language-level risks persist for Vyper-based contracts | ▲ 2 |
| 52 | | C+ | C- | Stablecoin | $1.3B | USDD relies on TRX as a primary reserve asset, creating correlated collateral risk — a severe TRX drawdown could impair the overcollateralization ratio below the 130% minimum despite the current 200%+ buffer. | ▼ 2 |
| 53 | | C | C+ | Stablecoin | $1.3B | Basis-trade yield strategy depends on persistent positive funding rates — prolonged negative funding can erode collateral backing | ▲ 2 |
| 54 | | B- | C+ | L1 | $1.2B | Bridge dependency — checkpoints to Ethereum create a trust assumption and potential attack vector; the PoS Bridge secures over $1B in locked assets with a validator multisig | ▼ 4 |
| 55 | | B | C- | Yield | $1.2B | Spark Savings (sDAI/sUSDS) depends entirely on the Sky (formerly Maker) DSR/SSR rate, which is governance-controlled. Rate changes (e.g., the March 2025 cut from 6.5% to 4.5% and the April 2026 cut to ~3.65%) cause rapid TVL swings as yield-seekers migrate, creating reflexive inflow/outflow dynamics. | → 0 |
| 56 | | B- | B+ | Yield | $1.2B | Ethena USDe concentration: a significant share of Pendle's $1.2B TVL remains in Ethena-linked pools; a USDe depeg or yield collapse would directly impact the majority of the deposit base | ▲ 5 |
| 57 | | C+ | C | DeFi | $1.2B | Smart Collateral and Smart Debt create reflexive leverage loops up to 39x theoretical max | ▲ 2 |
| 58 | | C+ | C+ | Bridge | $1.2B | Smart contract exploit in CommitStore or OnRamp/OffRamp contracts could allow forged cross-chain messages to trigger token minting without corresponding locks — the classic infinite-mint bridge exploit. The Risk Management Network (RMN) adds a second validation layer, but both layers must simultaneously fail for this to succeed. | — |
| 59 | | B- | B | Lending | $1.1B | 2024 governance attack extracted $24M COMP from treasury via coordinated whale voting (Proposal 247) | → 0 |
| 60 | | B- | D+ | Liquid Staking | $1.1B | Multi-LST Infinity pool aggregates risk from all supported LSTs; a single LST depeg can poison the entire pool through arbitrage-driven toxic asset accumulation | ▲ 4 |
| 61 | | C+ | C | RWA | $1.0B | Franklin Templeton (transfer agent) retains unilateral power to freeze, clawback, and restrict BENJI token transfers on all nine blockchains — tokens are not censorship-resistant | ▼ 1 |
| 62 | | B | B | Liquid Staking | $1.0B | 8 ETH minipool operators bear outsized slashing risk relative to their bond, with losses partially socialized to rETH holders | → 0 |
| 63 | | B- | C+ | Lending | $1.0B | Solana network risk concentration: Kamino Lend operates exclusively on Solana, which has experienced multiple network outages (2021-2023). A prolonged Solana outage during a market sell-off would prevent liquidations from occurring, allowing unhealthy positions to accumulate bad debt. | — |
| 64 | | C+ | C+ | Lending | $1.0B | Unified liquidity market allows risk spillover from one toxic asset to contaminate all lending positions | — |
| 65 | | C+ | C | Liquid Staking | $1.0B | BTC custody risk: Lorenzo holds custodied Bitcoin on behalf of stakers — a custody provider failure or hack would result in permanent BTC loss for stakers | ▲ 5 |
| 66 | | C | B- | Lending | $1.0B | Oracle manipulation on long-tail collateral remains Venus Core Pool's primary structural risk: the March 2026 Thena THE-token exploit caused $2.15M in bad debt despite the Resilient Oracle upgrade, as attackers with sufficient capital can still manipulate prices of illiquid assets through DEX spot price inflation before oracle bounds activate. | — |
| 67 | | C | C+ | Lending | $1.0B | History of severe incidents: $200M+ XVS price manipulation cascade (2021), $100M+ bad debt from BNB bridge hack (2022), and a March 15, 2026 donation attack extracting $3.7M via supply cap manipulation (attacker accumulated 12.2M THE tokens over 9 months to bypass supply limits) | → 0 |
| 68 | | B- | D | RWA | $964M | BCAP is a tokenized venture capital fund where the underlying portfolio consists of illiquid blockchain startup investments. NAV is determined by periodic fund valuations rather than real-time market pricing, creating potential for stale or inaccurate pricing between valuation events. | → 0 |
| 69 | | C | C+ | RWA | $911M | Tokenized equities depend on off-chain broker-dealer custody via Ondo's in-house SEC-registered subsidiary (formerly Oasis Pro Markets, acquired Oct 2025) — regulatory enforcement against Ondo's integrated entity or SEC reporting failure could freeze all token redemptions | → 0 |
| 70 | | D | D+ | Restaking | $905M | LayerZero bridge configuration allowed attacker to mint 116,500 rsETH (~$292M, 18% of supply) on April 18, 2026 with no corresponding ETH on source chain; largest DeFi exploit of 2026 | ▲ 19 |
| 71 | | C+ | B- | Lending | $896M | Rehypothecation in vaults creates cross-vault contagion risk despite initial 'zero contagion' marketing claims — Jupiter COO acknowledged in December 2025 that 'very limited' contagion risk exists | ▼ 1 |
| 72 | | B- | C+ | Yield | $865M | Multi-strategy vaults deploy capital across Aave, Curve, Morpho, and EigenLayer simultaneously; hidden correlations between strategies mean diversification benefits evaporate during systemic DeFi stress events | ▼ 5 |
| 73 | | B- | D | RWA | $864M | Anemoy relies on Chronicle Protocol's RWA Oracle for on-chain NAV reporting of its tokenized funds, creating a single oracle dependency for pricing accuracy across its $567M AUM. Chronicle's Proof of Asset framework provides cryptographic verification, but a sustained oracle failure could delay redemptions. | → 0 |
| 74 | | B- | C+ | DEX | $839M | Admin key compromise led to $4.4M exploit in Dec 2022, exposing centralised control over pool parameters | ▼ 3 |
| 75 | | B- | B+ | DEX | $836M | Permissionless hooks execute arbitrary code on every swap, enabling novel attack vectors with 36% of analyzed hooks found potentially vulnerable | → 0 |
| 76 | | B- | D- | RWA | $808M | Fully centralized operations — WisdomTree controls all minting, redemption, and transfer allowlisting with no on-chain governance | ▲ 2 |
| 77 | | B | C+ | DeFi | $795M | Chainlink Labs retains significant centralized control over network operations, including node operator selection and staking pool parameters, though the network has operated reliably for 7+ years under this model and a decentralization roadmap is in progress. | ▼ 5 |
| 78 | | B+ | B | DEX | $791M | Sandwich attacks exploit constant-product AMM with 90% of blocks vulnerable to front-running | ▼ 1 |
| 79 | | B- | C- | Liquid Staking | $776M | Kinetiq holds 82.5% market share in Hyperliquid liquid staking, creating single-point-of-failure concentration risk for the entire Hyperliquid staking ecosystem. | ▼ 2 |
| 80 | | B | D- | RWA | $769M | Centralized mint/redeem gating via allowlist means Superstate (as digital transfer agent) can freeze or deny redemptions at will | ▼ 2 |
| 81 | | B- | B | RWA | $754M | Permissioned access and custodian dependency on BNY Mellon create single points of failure outside on-chain control | ▲ 5 |
| 82 | | C+ | D- | Liquid Staking | $750M | Centralized custody: all staked SOL is managed by Binance validators, creating a single-entity dependency for ~$1.1B in assets | ▼ 2 |
| 83 | | B | B+ | Liquid Staking | $730M | JitoSOL's MEV tip distribution depends on >95% of Solana validators running the Jito client, creating systemic centralization risk for the network | ▲ 1 |
| 84 | | C+ | B- | Liquid Staking | $730M | Validator sandwich attacks extracted 30K-60K SOL/month despite bans — MEV redistribution incentivizes exploitation | → 0 |
| 85 | | B- | C | Liquid Staking | $721M | osETH overcollateralisation model means validators bear first-loss risk — slashing or poor performance directly erodes their position before osETH holders | ▲ 1 |
| 86 | | C+ | B | Derivatives | $706M | JLP holders are the counterparty to all perp traders — during trending markets, the pool can suffer significant directional losses | → 0 |
| 87 | | C+ | C- | DeFi | $700M | CeDeFi hybrid model depends on centralized custody (CEFFU/Binance) remaining solvent and accessible; LCTs (Liquidity Custody Tokens) become worthless if CeFi custodian fails, combining centralized custody risk with decentralized protocol exposure | ▼ 3 |
| 88 | | B- | C | L1 | $658M | Novel consensus — Snowball protocol is less battle-tested than traditional BFT or Nakamoto consensus | ▼ 1 |
| 89 | | B- | C | Lending | $620M | Extreme TVL growth (1,000% YTD to $4.5B across Lista DAO) means the lending markets are largely untested under sustained bearish conditions | ▼ 1 |
| 90 | | B- | C- | Lending | $620M | Systemic concentration risk: Lista DAO controls nearly 50% of BNB Chain's entire staking market with 12M+ BNB staked, creating a single point of failure for the chain's security and liquidity | → 0 |
| 91 | | B- | C | Liquid Staking | $594M | Institutional node operator concentration (Coinbase, Kraken, Figment, Blockdaemon, Staked) creates correlated regulatory risk; SEC enforcement against any operator could cascade to validator shutdowns and LsETH yield failure | ▼ 2 |
| 92 | | C+ | D+ | Liquid Staking | $585M | LBTC's 1:1 BTC backing depends entirely on Babylon's Bitcoin staking security; any slashing event or Babylon exploit directly depegs LBTC across all 15 integrated chains | ▼ 1 |
| 93 | | C | D+ | Restaking | $585M | LBTC depends on Babylon's nascent BTC staking infrastructure which has no proven slashing enforcement mechanism yet | → 0 |
| 94 | | C | C- | DeFi | $581M | Governance was compromised in May 2023 when an attacker used a malicious proposal with hidden SELFDESTRUCT/CREATE2 logic to grant themselves 1.2M votes, exceeding the legitimate 700K votes. The attacker later returned control, but the attack vector demonstrated that DAO proposal auditing is insufficient to prevent governance takeover. | ▲ 4 |
| 95 | | C | C- | L1 | $580M | Tether concentration risk: Plasma's entire value proposition depends on USDT/USDT0 as the dominant activity driver — Tether CEO Paolo Ardoino personally invested, tether.wallet selected Plasma as one of four supported chains, and the majority of TVL is USDT0 deposits. If Tether withdraws support, pauses USDT0 on Plasma, or blacklists the chain, the economic activity and TVL collapses. | — |
| 96 | | C+ | D+ | L2 | $567M | Optimism's sequencer remains fully centralized, operated solely by OP Labs with no decentralized fallback or concrete timeline for decentralization. Multiple sequencer outages occurred in 2025 (August and November), confirming this as a live operational risk rather than a theoretical concern. During downtime, users cannot submit transactions and must wait ~12 hours to force-include via L1. | ▼ 1 |
| 97 | | D+ | D+ | L1 | $557M | Aster Chain launched mainnet in March 2026 with no public specification of its ZK proving system, VM architecture, or consensus mechanism, and no L1-specific audit has been completed. The $298M in TVL sits on unverified infrastructure — a critical bug in the ZK circuit could allow fraudulent state transitions that drain user funds without detection. | ▼ 1 |
| 98 | | C- | C- | Bridge | $542M | Centralized admin keys allow contract owner to pause mints/burns, modify fees, and mint tokens arbitrarily — GoPlus flagged these as material risks | ▲ 3 |
| 99 | | B- | C | Liquid Staking | $540M | slisBNB commands ~50% of BNB Chain staking market share, creating unprecedented concentration risk for the chain's validator set and security model | ▼ 3 |
| 100 | | C | D | Stablecoin | $501M | USX experienced a severe depeg to $0.10 in December 2025 due to secondary market liquidity exhaustion on Solana DEXs, though the underlying collateral remained fully backed and the peg was restored within hours. | ▲ 2 |
