AI Transparency Disclosure

1. AI Attest uses AI in Audit Mode

When you run a compliance audit, AI Attest sends your uploaded documents to a large language model (LLM) for analysis. The LLM classifies your documents against regulation-specific requirements and assesses whether each required field has supporting evidence.

You are interacting with an AI system. The gap report, article coverage scores, and remediation recommendations produced by Audit Mode are generated with the assistance of AI. These outputs should be reviewed by a qualified professional before being relied upon for compliance decisions.

2. Build Mode does not use AI

Build Mode — where you fill in compliance templates, submit artifacts through the governed pipeline, and generate compliance reports — is entirely deterministic. No AI models are involved. Phase gates, dependency validation, cascade invalidation, approval checkpoints, and the cryptographic audit trail are all computed by fixed algorithms. Build Mode outputs are not AI-generated content.

3. Which content is AI-generated

4. AI model providers

AI Attest integrates the following general-purpose AI models via their hosted APIs. You choose your preferred provider and supply your own API key.

No user data is used for model training. All model providers are contractually bound by their respective API terms of service, which prohibit training on API inputs.

5. Machine-readable marking

AI-generated reports include metadata identifying the content as AI-assisted, including the model provider and model identifier used for the analysis. This metadata is embedded in the JSON export of all audit results and is available via the API.

6. Limitations

AI Attest is a compliance assistance tool. AI-generated outputs may contain errors, omissions, or misclassifications. Specifically:

• —Document classification may incorrectly map a document to the wrong artifact type
• —Field coverage assessment may report a field as covered when evidence is insufficient, or vice versa
• —Readiness scores are approximate and should not be treated as legal determinations
• —Remediation recommendations are informational and do not constitute legal advice

All AI-generated outputs should be reviewed by qualified compliance professionals or legal counsel before being used for regulatory submissions or business decisions.

7. Classification status

AI Attest has been self-classified as a non-high-risk AI system under the EU AI Act. It does not fall under any category listed in Annex III. This classification is subject to review upon changes to the product, regulatory guidance, or applicable harmonised standards. The full classification document is published at aiattest.io/classification.