AAK-AIACT50-003: Deep-fake or public-interest text generation with no artificiality disclosure
- Severity
- MEDIUM
- Category
- Legal Compliance (
legal-compliance) - Scanner
agent_audit_kit/scanners/eu_ai_act_art50.py- Auto-fixable
- No.
- SARIF name
EuAiActArt50NoDeepfakeDisclosure
What it detects
A declared capability to produce deep-fake image, audio or video content, or to publish text intended to inform the public on matters of public interest, with no disclosure that the content is artificially generated or manipulated. Article 50(4) of the EU AI Act, in force since 2026-08-02, places this on the deployer, and 50(5) requires the disclosure at the latest at the time of first exposure. The paragraph exempts artistic, creative, satirical and fictional work, and text that had human editorial review with someone holding editorial responsibility. Evidence toward the duty, not a determination that it applies.
How to fix
Disclose that the content is artificially generated or manipulated, at first exposure and in the content's own presentation rather than in terms of service. Where the work is artistic or satirical, Art. 50(4) still expects the existence of generated content to be disclosed, in a way that does not hamper enjoyment of the work.
To confirm the finding is gone, report only this rule:
agent-audit-kit scan . --rules AAK-AIACT50-003
Limitations
The 50(4) exemptions are assessed by the words a project uses about itself: a file describing satire, fiction or human editorial review is skipped entirely. That is deliberately generous, because the failure mode of the opposite bias is accusing a newsroom of an offence.
References
No external references are recorded for this rule.
All rules. This page is generated from the rule registry when the documentation is built; to change it, change the rule in agent_audit_kit/rules/builtin.py. To give the rule a hand-written write-up, add docs/rules/AAK-AIACT50-003.md: the page then shows that text, with the registry reference appended under it.