Built for the AI Act, by design.
Last updated 4 July 2026
The EU AI Act sets out to make high-stakes AI safe, accountable and transparent. Those are the same demands regulated financial institutions already live under, and they are the principles Brayns was built on from day one.
Where Brayns sits
Brayns runs a regulated firm's own compliance operation, inside that firm's environment, on that firm's own policies. A named, accountable person sits above the workflow at all times. They set the policy, review what the system escalates, and stay answerable to the regulator. Brayns does not make the final call in that person's place. It executes the firm's SOPs and hands up anything that needs judgement.
We assess Brayns as not a high-risk AI system under Annex III of the EU AI Act. It is an internal operational tool for a regulated entity, run under real human oversight. It does not independently decide outcomes for individuals. The firm stays the accountable deployer, and Brayns supports the obligations that come with that.
Designed around the Act's principles
- Human oversight
- People stay accountable above the workflow. Edge cases go to a person in real time, and a policy change goes live only after a human approves it. The system never runs unsupervised.
- Transparency and traceability
- Every decision Brayns makes is logged with the reasoning behind it and the checks it ran. The audit trail is captured as the work happens, not written up afterwards.
- Data governance
- Brayns runs inside your own environment on your own data. Nothing is pooled with other customers and nothing trains a shared model. What is yours stays yours.
- Accuracy and robustness
- Brayns runs on your own SOPs and risk policy, grounded in how your firm actually works. Its decisions are consistent, explainable and ready for examination.
- Purpose-built model
- Brayns runs on a model built for compliance, not a general model with tools bolted on top. That is what lets it stay grounded, auditable and defensible.
Governance
Our Head of Compliance owns how the platform lines up with the Act as it comes into force. We treat that the way we ask our customers to treat compliance: as an operating discipline we keep current, not a box ticked once.
Status & contact
The EU AI Act is coming into force in stages through 2026 and beyond. We follow its implementation and update our approach as the detail firms up. This page is informational and not legal advice.
Questions about our AI Act position? Email contact@brayns.ai