Artificial Intelligence
EU AI Office opens anonymous reporting channel for AI Act breaches
The AI Office has launched an anonymous reporting tool for harmful practices linked to general-purpose AI models and certain AI systems, with confidentiality safeguards and staged anti-retaliation protection.
2026-05-13 · 1 min
The European Commission AI Officehttps://digital-strategy.ec.europa.eu/en/policies/ai-act-whistleblower-tool has launched the AI Act Whistleblower Tool, a channel for people professionally connected to an AI model provider to report harmful practices by providers of general-purpose AI models and certain AI systems.
The tool accepts anonymous reports in any EU language. The AI Office page says supporting documents can be uploaded through a secure inbox, and that the inbox lets whistleblowers track progress and answer follow-up questions while remaining anonymous.
The source frames confidentiality as central to the channel. The AI Office says it commits to a high standard of confidentiality and has documented internal procedures intended to maximise protection of whistleblower identities. The page also explains the staged legal position under the Whistleblower Directive.
Legal protection against retaliation under the Whistleblower Directive will extend to reports concerning infringements of the AI Act from 2 August 2026. Until then, the page says confidentiality remains the primary means of protection. It also notes that some AI-related activities linked to product safety, consumer protection, privacy and personal data, or information security may already fall within the Directive scope.
The page positions whistleblowing as an early signal mechanism for potential violations that could endanger fundamental rights, health or public trust. It does not say the tool is a general customer complaints route, and it does not create a new filing obligation for every AI provider.
The customer relevance is therefore narrow but material. It creates an official EU channel for AI Act concerns, confirms anonymous submission and follow-up mechanics, and states the date when AI Act whistleblower retaliation protection extends under the Directive.
The source also matters because it distinguishes confidentiality from statutory anti-retaliation protection. Until 2 August 2026, confidentiality is the stated primary protection for AI Act whistleblowers on this page. From that date, reports concerning AI Act infringements are brought into the Whistleblower Directive protection described by the AI Office.
The official source supports the channel, the confidentiality posture, the anonymous inbox mechanics and the 2 August 2026 legal-protection date. It does not specify wider internal policy steps for firms, so the article stays with those source-backed mechanics rather than adding an operational playbook.