Compliance · EU AI Act
Know which AI you deploy, before you have to declare it.
European Union Artificial Intelligence Act
Most organizations meet the Act as a deployer rather than a provider: you did not build the model, you use it. Article 26 asks you to follow the provider's instructions, put a named human in the oversight seat, and keep records of what you run. The transparency duties have applied since 2 August 2026, obligations for general-purpose AI models since 2 August 2025, and the high-risk deadlines now run to December 2027 and August 2028. All of them start from a register of the AI actually in use, and that is what KeyCaliber produces from read-only connections to the tools you already have.
What EU AI Act expects.
- A record of the AI systems you use and what each is for
- Human oversight assigned to a named person
- Use in line with the provider's instructions, with records kept
- Disclosure where people interact with AI or see AI-generated content
What KeyCaliber gives you.
The register, built from evidence
Every AI system in use — hosted services, AI inside apps you already approved, model runtimes on laptops and servers, and remote endpoints — with the signal that found it and when.
Accountability on the record
A named owner on every AI and SaaS app, the decision taken, who took it, when, and the note behind it.
AI switched on for you
Which approved products ship AI features and whether the vendor turns them on by default — the disclosure question an app list never answers.
The classification stays yours
The catalog offers a suggested risk class, labelled as a suggestion. Your classification is the one that stands. We answer clauses; we do not certify anyone.
Build the register before you are asked for it.
Connect the tools you already run and see your own picture.
Request a demo