Terms of use

Last updated: 15 September 2026

1. What Proofroom is

Proofroom is an evidence registry for AI agents. Operators register use cases ("rooms"), record hash-chained evidence events, and publish a public trust page. The registry reflects evidence submitted to it. It does not certify that an agent is accurate, safe, compliant, unbiased, or fit for any purpose.

2. No endorsement

Status labels, badges, chain checks, and "evidenced" / "proofed" language describe the presence and integrity of a submitted trail. They are not an endorsement of the agent, the operator, or any outcome the agent produces.

3. Your responsibilities

You are responsible for the accuracy of events you submit, for claim review of rooms you operate, for API key custody, and for complying with applicable law. Do not submit unlawful content, secrets you are not authorised to share, or evidence intended to mislead buyers.

4. Availability and correct-forward integrity

Evidence recording is best-effort and must not sit in your agent's critical path. Hash chains are append-only: we do not rewrite historical event hashes. Explained discontinuities may remain visible. Service availability is not guaranteed.

5. Free tier and fair use

Registration may be free. Claim review, rate limits, per-operator room limits, and auto-expiry of unclaimed drafts apply so the registry is not flooded with junk rooms. We may refuse, retire, or expire rooms that violate these terms or harm the registry.

6. Contact

Support: support@proofroom.ai. Security: security@proofroom.ai. See also Privacy and Imprint.