USD AI

USD AI receives a C risk grade (44/100) from Hindenrank, where lower scores indicate lower risk. Moderate risk — novel GPU-backed lending with institutional-grade Munich Re loss-given-default insurance substantially reducing depositor tail risk, offset by single-borrower concentration (Sharon AI $500M) and governance token (CHIP) instability. USD AI is a DeFi lending protocol that converts stablecoin deposits into GPU-backed loans for AI companies, using NVIDIA hardware as collateral through its CALIBER tokenization framework. The $CHIP governance token launched in April 2026, transitioning the protocol to DAO governance with CHIP stakers serving as first-loss insurance for depositors. A Munich Re-reinsured insurance policy (via Barker) now provides 100% loss-given-default coverage on GPU-backed loans, substantially strengthening the depositor protection layer. With $36.8M in total funding, Wilmington Trust as formal escrow agent, and a $500M Sharon AI facility among $1.2B+ in approved GPU lending facilities, the institutional foundation is substantial — but the C+ risk grade reflects continued novel GPU collateral mechanism risk and concentration in single large borrowers.

The key risks identified for USD AI are: (1) GPU hardware used as loan collateral depreciates quickly as new chip generations are released. Unlike crypto collateral that can be instantly liquidated, selling physical GPUs takes time. Munich Re's institutional insurance covers loss-given-default, but insurance claims take time to process and policy exclusions remain a tail risk. (2) The $500M Sharon AI facility is the largest known single-borrower exposure. A default or dispute from this counterparty could strain the loan book disproportionately, even with Munich Re coverage providing a financial backstop. (3) 80% of CHIP supply remains locked to insiders and VCs following the April 2026 TGE, with first major insider unlocks beginning April 2027. CHIP is a governance-only token with no revenue rights, trading at -64% from its ATH; future unlock events could create sustained sell pressure.

USD AI scores 44/100 across eight risk dimensions: Mechanism Novelty: 9/15, Interaction Severity: 7/20, Oracle Surface: 5/10, Documentation Gaps: 3/10, Track Record: 3/15, Scale Exposure: 5/10, Regulatory Risk: 5/10, Vitality Risk: 7/10. The highest risk area is Vitality Risk at 7/10.

Among 77 rated RWA protocols on Hindenrank, USD AI ranks #56 by safety (lowest risk score = safest). Its 44/100 risk score and C grade place it among the riskier RWA protocols.

USD AI scores 3/15 on the Track Record risk dimension, indicating some history of security incidents or exploits. Higher scores reflect more severe or frequent incidents. Review the full risk report for details.