Trusted AI evaluation in TEEs

We come as a technology provider, offering a hybrid approach that leverages Trusted Execution Environments (TEEs) and blockchain infrastructure to enable secure, trusted, confidential, and decentralised AI processing.

Alignment with the call

• TEE‑based assessments: Running GPAI performance and risk tests entirely inside Trusted Execution Environments preserves model and data confidentiality and integrity, delivering the “reliability” and “robustness” required by the topic.

• Blockchain traceability: Anchoring scores and audit reports on an immutable ledger yields verifiable proof of compliance, fully aligned with the forthcoming AI Act and the AI Office’s supervisory role.

• Data monetisation & reputation: A reputation‑driven reward system pays data providers for high‑quality test suites, creating strong economic incentives for trustworthy, diverse benchmarks.

• Targeted impact: By combining security (TEEs), auditability (blockchain) and sustainable economics (incentives), the solution accelerates adoption of evaluation methodologies by developers, regulators and policy‑makers—directly meeting the expected outcomes of the call.

We are actively seeking a consortium to co-develop and demonstrate this vision. If you're interested in building a trusted AI assessment, we’d be glad to connect.