Risk Labs Comprehensive Report

UMA Protocol | Across Protocol | Corporate Overview

Published: March 16, 2026 | Research conducted by multi-agent DeFi analysis team with independent QA verification A synthesis report written by the apriori-writer agent.

tl;dr

• Risk Labs is a Cayman Islands nonprofit behind two DeFi protocols: UMA (optimistic oracle) and Across (intents-based cross-chain bridge). Founded by Goldman Sachs alumni Hart Lambur and Allison Lu; $51M raised total with Paradigm as lead investor.

• UMA powers Polymarket’s market resolution but has suffered high-profile oracle failures — including a $7M governance attack by a whale who accumulated 25% of voting power, and a $237M market where value at stake exceeded UMA’s entire market cap. The response (UMIP-189/MOOV2) traded decentralization for reliability by whitelisting 37 proposer addresses.

• UMA’s AI initiative — the Optimistic Truth Bot — runs in shadow mode at 95% accuracy on binary markets but does not address the whale-dominated dispute voting layer.

• Across has processed $35B+ in lifetime volume with zero exploits and co-authored ERC-7683 (cross-chain intent standard adopted by 35+ L2s), but captures zero protocol revenue.

• The Bridge Across — the most significant current development — is a proposal to convert the Across DAO into a U.S. C-corporation (AcrossCo), offering token holders 1:1 equity conversion or a USDC buyout at $0.04375 (25% premium). Snapshot vote scheduled for March 26, 2026.

• Governance concerns: Unresolved allegations of $23M in ACX tokens directed from the DAO treasury to Risk Labs via governance proposals where team wallets may have cast deciding votes — adding context to the corporate restructuring debate.

Table of Contents

1. Risk Labs: Company Overview
2. UMA Protocol & Polymarket
3. Across Protocol: Data & Competitive Position
4. The ACX-to-Equity Conversion Proposal
5. Key Risks & Open Questions
6. Timeline of Key Events
7. Sources

1. Risk Labs: Company Overview

Founding & Leadership

Risk Labs Foundation is a Cayman Islands-registered nonprofit foundation that serves as the core development and operational entity behind two DeFi protocols: UMA and Across.

• Co-Founders: Hart Lambur (CEO) and Allison Lu
• Hart Lambur: Columbia CS (Magna Cum Laude, 2005). Interest rate trader at Goldman Sachs (2005–2013), lived through the 2008 crisis on the trading desk. Co-founded Openfolio (personal finance platform, acquired 2017). Founded UMA in 2018.
• Allison Lu: Economics & Management, MIT. VP at Goldman Sachs for six years, managed a $30B balance sheet and $100M+ in annual revenue. Elevated desk to #1 in U.S. Treasuries (primary and secondary). Has since stepped back from day-to-day operations; involved with Founders Pledge.
• Nick Pai: Tech Lead, Across Protocol. Joined Risk Labs in 2020.
• Kevin Chan: Project Lead, Across Protocol. Central figure in the DAO treasury controversy (see Section 5).

Structure: Remote-first organization. Cayman Islands nonprofit. Lean leadership team with a “Chief of Staff to Founder” role in job postings. Exact headcount not publicly disclosed.

Funding History

Strategic Architecture: The Two-Protocol Flywheel

Risk Labs operates a vertically integrated infrastructure stack:

1. UMA (oracle layer): Provides trust and verification infrastructure. The optimistic oracle is a general-purpose system for asserting and verifying data on-chain — the “source of truth” layer.
2. Across (application layer): Uses UMA’s oracle as its security backbone for cross-chain bridging. Across is the highest-profile commercial application built on UMA’s infrastructure.

The flywheel: Across drives demand for UMA’s oracle services, and UMA’s security model gives Across a differentiated trust architecture vs. competing bridges.

Strategic Evolution

2. UMA Protocol & Polymarket

How UMA’s Optimistic Oracle Works

UMA’s oracle operates on two layers:

Layer 1 — The Optimistic Oracle (OO):

1. Request: A data consumer (e.g., Polymarket) requests information
2. Propose: A proposer submits an answer + posts a bond as collateral
3. Challenge Period: Opens (typically 2 hours for Polymarket). If unchallenged, the answer is accepted as truth.
4. Dispute: If challenged, the disputer posts their own bond and it escalates to Layer 2.

Layer 2 — The Data Verification Mechanism (DVM):

• UMA token holders vote on the correct answer over 48–96 hours
• Token-weighted voting: voters who vote with the majority are rewarded; incorrect/absent voters are slashed
• This is the ultimate backstop for all UMA-secured protocols

DVM 2.0 Staking Economics:

• Staking required to vote (minimum raised from 500 to 1,000 UMA in Nov 2025)
• Emissions: 0.05% of UMA supply per voting round to active voters
• Estimated staking APY: ~30% for consistent voters
• 7-day unstake cooldown
• Slashing redistributes tokens from inactive/incorrect voters to correct voters

Polymarket Integration: The Exact Mechanism

Polymarket uses a UMA CTF (Conditional Token Framework) Adapter — a smart contract sitting between Polymarket’s markets and UMA’s oracle:

1. Market hits resolution conditions → Adapter sends price request to UMA’s OO
2. Proposer submits answer (Yes/No) with bond
3. 2-hour liveness window opens for disputes
4. Critical design feature — Two-round dispute mechanism: The first dispute is ignored and a new identical request is created (prevents malicious single disputes from slowing resolution). Only if the second request is also disputed does it escalate to full DVM vote.
5. DVM token-holder vote (if escalated) → 48–96 hours → majority determines outcome
6. Finalization → Polymarket conditional tokens settle accordingly

Polymarket actually uses three oracle systems:

• UMA Protocol — most markets, especially subjective judgment calls
• Chainlink — markets with clearly defined, objective data feeds
• Markets Team — Polymarket’s internal team for certain market types

Major Controversies & Oracle Failures

1. The Barron Trump DJT Meme Coin Override (June 2024)

• Market: “Was Barron Trump involved in creating the $DJT token?”
• UMA’s DVM repeatedly resolved: “No”
• Polymarket’s response: Overrode the oracle, stating it was “conclusive” that Barron Trump was “in fact, involved in some way”
• Significance: First public demonstration that Polymarket retains centralized override power over its “decentralized” oracle. Raises fundamental questions about actual decentralization.

2. The $7M Ukraine Mineral Deal Governance Attack (March 2025)

The single most significant oracle failure in UMA/Polymarket history:

• Market: “Will Ukraine agree to Trump’s mineral deal before April?”
• The attacker: An Ethereum wallet identified as BornTooLate.eth accumulated ~1.3 million UMA tokens over the preceding year (~$2M+ cost), becoming a top-5 governance staker
• The attack: Cast ~5 million UMA tokens across three accounts (25% of total votes) to force a “Yes” resolution
• The result: Market moved from 9% → 100% despite no official agreement existing
• Polymarket’s response: Confirmed the oracle reached the “incorrect outcome” but called it “unprecedented” and issued no refunds (“this wasn’t a market failure”)
• Financial impact: Largest winner gained ~$55,000; largest loser lost ~$73,000
• Core issue exposed: 95% of UMA tokens held by large holders (per IntoTheBlock). Token-weighted voting fundamentally favors capital over truth.

Verified across: CoinDesk, The Defiant, CoinTelegraph, Yahoo Finance, The Block, CoinMarketCap

3. The Zelenskyy Suit Controversy (July 2025)

• Market: “Will Zelenskyy wear a suit before July?” — $237M in volume
• What happened: On June 24, Zelenskyy appeared at a NATO event wearing a black jacket, matching trousers, and a collared shirt. BBC and New York Post described it as a suit.
• Resolution: Initially proposed as “Yes,” then over nine days of disputes and re-proposals, flipped to “No” by UMA whale voters
• The whale problem (quantified): Top 10 UMA voters held ~6.5M UMA tokens (~30% of average vote participation). Allegations emerged that large UMA holders also held positions on the Polymarket market itself — a direct conflict of interest.
• THE KEY INSIGHT — Scale Mismatch: UMA’s total market cap was ~$95M. This single Polymarket market had $237M in volume. The economic value at stake exceeded the entire market cap of the oracle securing it — a fundamental security model failure.

4. The UFO Declassification Market

• Polymarket resolved “YES” on a $16M market asking whether Trump would declassify UFO files in 2025, despite no documents being released
• Users labeled the outcome a “scam” and criticized the “proof-of-whales” model

5. Other Notable Disputes

• Venezuela’s contested election — Criticized for subjective decision-making
• OceanGate submarine — Disputed over interpretation of “finding” wreckage vs. the vessel

Systemic Response: UMIP-189 and the MOOV2 Upgrade (August 2025)

In response to the governance attacks, UMA passed UMIP-189 on August 6, 2025:

The tradeoff: MOOV2 trades decentralization for reliability. Critics call it centralization of the oracle around a small pre-approved group. Supporters argue it’s pragmatic given demonstrated vulnerability.

The Fundamental Security Model Problem

UMA’s security relies on a game-theoretic assumption: the cost of corrupting the oracle should exceed the profit from corruption. This is the “cost-of-corruption” model.

Why it’s broken at scale:

• When Polymarket markets grow large enough, the profit from manipulating a resolution can exceed the cost of acquiring enough UMA tokens to control the vote
• The Zelenskyy market ($237M volume) vs. UMA market cap (~$95M) is the clearest demonstration
• MOOV2 fixes the proposal side but does not fix the dispute-resolution voting problem
• Just four whales reportedly control over 40% of UMA supply; the largest single whale holds ~25%

UMA’s AI Oracle Initiative: The Optimistic Truth Bot

UMA launched the Optimistic Truth Bot (OTB) in March 2025 as an experimental AI-powered proposer for the Optimistic Oracle. As of March 2026, the bot runs in shadow mode — processing real Polymarket markets in parallel with human proposers but not submitting proposals on-chain. It publishes recommendations publicly on X at @OOTruthBot.

Architecture: Mixture of Experts

The OTB uses a three-layer design:

UMA’s blog references GPT-4, Claude 3, and Llama 4 as enabling models, though the specific model(s) powering the current deployment are not confirmed. Perplexity’s API is explicitly used for the web-search solver.

Accuracy Metrics

For comparison, human proposers achieve a 98.2% assertion acceptance rate with only 0.4% facing genuine accuracy disputes.

Current Status

The OTB is not yet making on-chain proposals. UMA’s plan is to integrate it once accuracy is sufficiently high, subject to the same dispute process as human proposers. The vision: “AI proposes, humans decide” — AI handles routine cases, human oversight remains the backstop.

Critical Limitation: AI Does NOT Solve the Whale Problem

The OTB addresses the proposal layer, NOT the voting/dispute layer. If a proposal is disputed, it still goes to UMA token-weighted voting — where the whale concentration problem exists. AI-based proposals make the system faster and cheaper, but do not reform the dispute resolution mechanism that has been exploited.

EigenLayer Next-Gen Oracle (Research Phase)

The more promising initiative for the whale problem is UMA’s collaboration with Polymarket and EigenLayer on a next-generation oracle:

• Multi-token dispute resolution (not just UMA tokens)
• EigenLayer’s intersubjective security model to mitigate economic bribery attacks
• Prediction markets could stake their own native assets instead of relying solely on UMA
• Dynamic bonding system conducive to AI agent participation
• Could dilute whale influence by diversifying the voting base beyond UMA holders

This is still in the research phase — no deployed system yet. But it represents the most credible path to solving the structural security model failure.

Connection to UMIP-189 / Managed Proposers

The AI oracle and managed proposers (UMIP-189) are currently parallel tracks:

• Managed proposers: Human-curated whitelist addressing proposal quality (177 addresses as of Nov 2025, up from initial 37)
• OTB: AI-assisted proposals as separate R&D effort
• Logical convergence: If the OTB achieves sufficient accuracy, it could become a whitelisted managed proposer — but this hasn’t happened yet

Existential Threat: The POLY Token

Polymarket has teased a POLY token (reported October 2025) that could potentially replace UMA for dispute resolution. No tokenomics have been disclosed. If Polymarket internalizes its oracle function, UMA loses its most important use case.

UMA Token Metrics (March 2026, approximate)

Data caveat: Price data shows significant discrepancy between aggregators. TVL data is also inconsistent ($1.5M vs. $18M across sources). These figures should be verified on live dashboards.

3. Across Protocol: Data & Competitive Position

How Across Works (Technical Detail)

Across’s intents-based architecture decouples user experience from backend settlement:

1. User submits intent: Calls deposit on the SpokePool contract on the origin chain, escrowing assets with metadata (recipient, destination chain, relayer fee)
2. Relayer competition: Bonded relayers compete to fill the intent on the destination chain using their own capital → user receives funds in seconds to <1 minute
3. Bundling: A “dataworker” aggregates fulfilled intents into repayment bundles
4. Optimistic settlement: Bundles submitted to UMA’s oracle on Ethereum mainnet with a challenge period
5. Dispute resolution: If challenged, UMA token holders vote. Incorrect proposers are slashed.

Key difference from competitors: No multisigs, no centralized validators, no lock-and-mint. The optimistic oracle model is architecturally distinct from the validator/multisig approaches that have been exploited for billions (Ronin, Wormhole pre-upgrade, Nomad).

Security Track Record

• 18 comprehensive audits (primarily OpenZeppelin)
• 232 issues identified and addressed across all audits
• Zero security exploits in production — ever
• No multisigs or centralized validators

Supported Chains (15+)

Ethereum, Arbitrum, Optimism, Base, zkSync, Blast, Scroll, Zora, World Chain, Linea, Polygon, BNB Chain, Solana, Hyperliquid, and others being onboarded.

Smaller coverage than Stargate (~80 chains) or Wormhole (~30+), but focused on high-volume Ethereum + L2 corridors.

Bridge Data & Metrics

Critical distinction: Across generates fees from bridge transactions, but these fees flow to relayers and liquidity providers. The protocol itself captures zero revenue. Fee revenue ≠ protocol revenue. This is a key motivation behind the corporate restructuring.

Data caveat: The $35B lifetime volume figure comes from protocol and partner marketing materials. The most recent independently verifiable data point found was $20.7B at an earlier date. DeFiLlama (defillama.com/bridge/across) is the authoritative neutral source.

Competitive Landscape (Detailed)

Note: Stargate/LayerZero merger in 2025 ($110–120M acquisition) creates a formidable integrated competitor.

Where Across Wins

1. Ethereum + L2 routes: Consistently rated fastest and cheapest for transfers under $1,000 (often <$1 per ETH bridged)
2. Institutional distribution: Uniswap and MetaMask integrations give unmatched retail access — every Uniswap bridge transaction flows through Across
3. Standards ownership: Co-authored ERC-7683 with Uniswap (see below)
4. Security architecture: Zero exploits, no multisig attack surface

Where Across Loses

1. Chain coverage: 15+ vs. 80+ (Stargate) — not competitive for non-EVM bridging beyond Solana
2. Speed on some routes: deBridge and Stargate V2 achieve sub-2-second on some routes
3. Fee competitiveness: Synapse reportedly offered lower fees on 45/60 tested routes in one benchmark
4. Revenue capture: $0 protocol revenue signals no fee switch activated

Key Partnerships & Integrations

Uniswap (Flagship)

• Across powers in-app bridging directly within the Uniswap interface
• Supported chains: Ethereum, Base, Arbitrum, Polygon, OP Mainnet, Zora, Blast, World Chain, zkSync
• Co-authored ERC-7683 (cross-chain intent standard) with Uniswap Labs

ERC-7683: The Strategic Moat

• Universal framework for intent-based cross-chain systems
• 35+ L2s and sidechains have adopted/expressed support (including Arbitrum, Optimism, Base)
• As the standard gains adoption, Across’s relayer network becomes the default fulfillment layer
• This positions Across not just as a bridge, but as core infrastructure for cross-chain intent settlement

MetaMask

• Integrated into MetaMask for cross-chain bridging (massive user base access)

PancakeSwap

• Cross-chain swaps powered by Across via ERC-7683 (June 2025)
• Supports BNB Chain ↔ Arbitrum ↔ Base in a single transaction
• Combines bridging + swapping into one action

Other: ZeroDev (account abstraction), LI.FI (aggregation), various bridge aggregators

ACX Token Metrics (March 2026, approximate)

Token Distribution:

4. The ACX-to-Equity Conversion Proposal

“The Bridge Across” — The Most Significant Development at Risk Labs

Proposal Name: “The Bridge Across” Posted: March 11, 2026, on the Across governance forum (forum.across.to/t/the-bridge-across/2097) Proposed by: Risk Labs

What Is Being Proposed

Risk Labs proposes converting Across from a DAO + token governance structure to a U.S. C-corporation called “AcrossCo.” The new entity would:

• Take over development, partnerships, and commercialization
• Hold the protocol’s intellectual property
• Enter enforceable contracts and structured revenue agreements
• The underlying bridge infrastructure would remain open and permissionless

Exact Terms (verified across 5+ independent sources)

Option A: Equity Exchange

• Convert ACX tokens to equity in AcrossCo at 1:1 token-to-share ratio
• Holders with >5 million ACX can convert directly to equity
• Smaller holders (minimum 250,000 ACX, ~$10,000) participate through a no-fee Special Purpose Vehicle (SPV)

Option B: USDC Buyout

• Sell tokens for USDC at $0.04375 per token
• Represents a 25% premium over the trailing 30-day average price
• Protocol’s liquid assets (~equivalent to market cap) finance the buyout
• Six-month redemption window opens within three months of proposal passing

Governance Timeline

CRITICAL: As of March 16, 2026, the binding Snapshot vote has NOT yet taken place. The proposal is in discussion/temperature-check phase.

Risk Labs’ Stated Rationale (Direct Quotes)

“As Across deepens our work with institutional and enterprise partners, the token and DAO structure has materially impacted our ability to close partnerships and integrations.”

“Enterprise partners need enforceable contracts. Revenue agreements need a legal counterparty. The kinds of deals that would drive the next phase of growth require a structure that a DAO, today, simply can’t provide.”

“Transitioning to a traditional legal entity would meaningfully improve our ability to enter enforceable contracts, structure revenue agreements, and deliver more value to Across stakeholders.”

Market & Community Reaction

• Price: ACX surged 80–85% within 24 hours; trading volume hit ~3.5x market cap
• Market cap: Briefly reached ~$45M
• Sentiment: Market is pricing this as value-accretive
• The 25% premium buyout provides a price floor for dissenters
• Community discussion ongoing — formal vote hasn’t happened yet

Supportive Arguments

• Provides legal clarity for institutional partnerships
• Offers tangible equity stake with potential for dividends, M&A, or IPO
• Addresses real limitations of DAO structures for commercial operations
• Market price reaction strongly positive

Critical Arguments

• Philosophical retreat from decentralization
• One of the first major protocols to argue DAOs are inferior to traditional corporate structures
• Equity in a private company is illiquid vs. freely tradeable token
• SPV structure introduces intermediation for smaller holders
• Regulatory implications of token-to-equity conversion are untested
• Critics could argue Risk Labs is dissolving the DAO partly to escape governance accountability (see $23M controversy below)

Legal & Regulatory Implications

• Converting governance token to C-corp equity explicitly makes the asset a security under U.S. law
• Token-to-equity swap and USDC buyout both likely trigger taxable events
• Direct equity conversion for 5M+ ACX holders may require accredited investor verification
• Paradigm’s support (as $41M lead investor) suggests comfort with the regulatory path
• This is precedent-setting — how regulators respond could create a template for or against similar future conversions

Broader Industry Implications

1. DAO disillusionment signal: If successful, sets precedent that DAOs aren’t the default organizational structure
2. Token-to-equity template: Creates a mechanism for how protocols might “re-incorporate”
3. Regulatory pragmatism: Acknowledges that operating in a regulatory gray zone as a DAO may be costlier than incorporating traditionally
4. VC alignment: Paradigm would gain traditional equity rights (board seats, liquidation preferences) instead of governance tokens with unclear legal standing

5. Key Risks & Open Questions

The $23 Million DAO Treasury Scandal

This is a critical piece of context for evaluating Risk Labs’ governance credibility:

Allegations (surfaced 2025, raised by Ogle, founder of Glue):

• Kevin Chan (Across project lead) submitted a governance proposal in October 2023 to transfer 100M ACX tokens (~$15M) from the DAO treasury to Risk Labs
• On-chain analysis allegedly showed Kevin used a second wallet (“maxodds.eth”) to secretly cast “yes” votes on his own proposal
• Without the team’s own votes, the proposal allegedly would not have passed
• A second “retrospective funding” proposal for 50M ACX (~$7.5M) followed. “maxodds.eth” and a new wallet funded by it allegedly contributed 44% of “yes” votes
• Total allegedly directed to Risk Labs: ~$23M in ACX tokens
• No public audit or transparent accounting of how funds were used

Risk Labs’ Response (Hart Lambur):

• Called allegations “completely untrue”
• Stated Risk Labs is a nonprofit Cayman foundation with fiduciary responsibilities
• Claimed team members purchased tokens independently and voted openly
• Stated all wallet addresses are publicly disclosed
• Asserted funds used for protocol development and team expansion
• Questioned accuser’s credibility (ties to competing projects)

Status: Unresolved. No legal proceedings filed. Absence of transparent accounting remains a legitimate concern.

Relevance to equity conversion: Critics argue Risk Labs may be dissolving the DAO partly to escape governance accountability. Supporters argue a C-corp with board oversight and auditing would actually improve accountability.

For UMA Protocol

For Across Protocol

For Risk Labs Corporate

Open Questions to Watch

1. March 26 vote: Does the community approve DAO → C-corp conversion?
2. SEC response: How do regulators view the token-to-equity conversion?
3. Institutional deals: Does corporatization actually unlock enterprise partnerships?
4. POLY token: Does Polymarket internalize its oracle, killing UMA’s primary use case?
5. UMA security model: Can it be redesigned to handle markets exceeding its market cap?
6. Precedent effect: Do other protocols follow Across’s lead on DAO → C-corp?
7. $23M accountability: Will the treasury controversy be transparently resolved?

6. Timeline of Key Events

7. Sources

All findings were cross-verified by an independent QA supervisor agent. Full source list below.

UMA / Polymarket Sources

• UMA Docs: How does UMA’s Oracle work?
• UMA Docs: DVM 2.0
• UMA Blog: What is UMA’s Optimistic Oracle?
• UMA Blog: Improving Oracle Efficiency with Managed Proposers
• UMA Discourse: UMIP-189
• Polymarket Docs: Resolution via UMA
• Polymarket Legacy Docs: Polymarket + UMA
• GitHub: Polymarket UMA CTF Adapter
• CoinDesk: Polymarket Suffers UMA Governance Attack
• CoinDesk: $160M Controversy Over Zelenskyy Suit
• CoinDesk: Polymarket’s POLY Could Bring Oracle’s Home
• The Block: Governance Attack ‘Unprecedented’
• The Block: UMA Oracle Update to Limit Proposals
• The Block: Polymarket Contradicts UMA on Barron Trump
• The Defiant: $7M Ukraine Mineral Deal Debacle
• The Defiant: Zelenskyy Suit Market Resolves
• Decrypt: Polymarket Rules ‘No’ on $237M Bet
• Yahoo Finance: ‘This Isn’t Decentralized’
• Webopedia: Polymarket’s UMA Controversial?
• Orochi Network: Oracle Manipulation in Polymarket 2025
• Cryptopolitan: Community Protests Oracle Vote
• CryptoSlate: Whales Forced UFO Vote

UMA AI Oracle Sources

• UMA’s AI Experiment: Can AI Agents Enhance the Optimistic Oracle?
• Inside UMA’s Optimistic Truth Bot
• AI Is Helping Us Find the Truth — UMA Blog
• Managed Proposers Update — UMA Blog
• UMA, Polymarket, and EigenLayer Research a Next-Gen Prediction Market Oracle
• UMA’s Truth Bot Joins Polymarket, Accuracy Hits 95% — Stocktwits
• AI Agents Are Quietly Rewriting Prediction Market Trading — CoinDesk

Across Protocol Sources

• Across Protocol Official
• Across Documentation
• Across V3: First Intents-Based Protocol
• The Bridge Across — Governance Forum
• CoinDesk: ACX Rockets 80%
• The Block: ACX-to-Equity Exchange
• Crypto.news: ACX Jumps 85%
• The Defiant: DAO to Private Company
• Crypto.news: Token-to-Equity Shift
• DL News: DAO-to-Company Proposal
• Odaily: Across Leads the “Rebellion”
• Across Raises $41M — Official Blog
• The Block: Paradigm Leads $41M Round
• CryptoNews: Across Raises $41M
• Uniswap Labs Integrates Across
• ERC-7683 Standard Proposal
• PancakeSwap Cross-Chain via Across
• ERC-7683 Official
• OpenZeppelin: How Across Secured $30B+
• Across: Why We’ve Never Been Hacked
• UMA: Case Study Securing Across
• DeFiLlama: Across Protocol
• DeFiLlama: Bridge Rankings
• Dune: Across Bridge Stats

$23M Treasury Scandal Sources

• Protos: Across Accused of Looting DAO Treasury
• CoinTelegraph: Founders Accused of Funneling $23M
• CryptoPotato: Co-Founder Responds
• The Block: ACX Drops 10% Amid Allegations
• Bitget News: Governance Scandal

Risk Labs Corporate Sources

• Risk Labs Foundation
• Risk Labs — Crunchbase
• Hart Lambur — IQ.wiki
• Allison Lu — IQ.wiki
• UMA 2024 Year in Review
• CoinGecko: UMA
• CoinMarketCap: UMA
• CoinGecko: Across Protocol
• CoinMarketCap: Across Protocol

QA Verification Notes

This report was produced by a 4-agent research team:

• 3 DeFi Market Analyzer agents conducted parallel research on UMA/Polymarket, Across Protocol, and Risk Labs corporate
• 1 DeFi Agent Supervisor independently verified all findings and flagged potential fabrication risks

Confidence Levels:

• HIGH confidence: Founding team, funding history, ACX equity proposal terms ($0.04375 buyout, 1:1 equity, 5M threshold, SPV structure, March 26 vote), UMA governance attack details (BornTooLate.eth, 5M UMA, March 2025), UMIP-189/MOOV2 changes (37 addresses, August 6, 2025), $23M treasury allegations, Zelenskyy suit $237M market, security audit record (18 audits, zero exploits), OTB accuracy metrics — all verified across 3+ independent sources
• MEDIUM confidence: TVL figures (~$32–52M range across sources), lifetime volume ($19B–$35B range depending on date), UMA token price (~$0.42–0.68 discrepancy), ACX circulating supply (~658–702M discrepancy)
• DATA GAPS (explicitly noted): Current daily bridge volumes, precise market share %, specific fee revenue breakdowns, Risk Labs current valuation, exact employee count, Risk Labs treasury balance, detailed fund deployment accounting for $23M controversy

No data was fabricated in this report. Where specific numbers could not be independently verified or showed discrepancies across sources, this is explicitly noted with ranges and caveats.