← Vetro Attest · Free Art. 12/14 checker · Free EU AI Act tools

EU AI Act Article 5: prohibited practices and how to show an agent avoids them

Article 5 lists AI practices that are prohibited outright, such as manipulation that causes significant harm, exploiting vulnerabilities, social scoring, and certain biometric uses. Unlike the high-risk rules, there is no compliance route: the practice is not allowed.

Why agents need a technical control

An LLM agent with tools can drift into behaviour nobody designed. Prohibition is best enforced before the action runs: a policy that denies categories of tool calls or content and fails closed.

What to record

Every blocked or warned action with the rule that fired and the reason. A tamper-evident record of refusals is evidence the control was live, not only documented.

Gaps to state honestly

A policy catches what you can express as a rule. Whether a use case falls under Article 5 is a legal assessment; the log only shows what the system did.

How Vetro helps

Vetro writes every model call, tool call and policy decision to a hash-chained, ed25519-signed ledger that anyone can verify with the public key. Vetro Attest verifies a ledger or OTLP GenAI trace for free and, for 49 USD, issues a counter-signed report mapping the evidence to the EU AI Act, including an explicit list of what the log cannot show. This is technical evidence, not legal advice.

See a sample ledger verified in one click, check your own logs free or get the countersigned EU AI Act report.