EU AI Office launches whistleblower tool to report AI Act breaches

In the FAQs, the EU AI Office recognises that those who are professionally connected to AI model providers and certain AI systems (such as current or former employees, self-employed collaborators, shareholders, or management) may be particularly well placed to support early detection of breaches. The tool intends to provide a secure and accessible means for whistleblowers to report suspected breaches of the AI Act anonymously. The tool includes a secure inbox that allows the reporters to upload supporting files, receive updates on the status, answer any follow-up questions and receive feedback on the outcome of the report. The EU AI Office assures whistleblowers that any report will be kept confidential and secure via encryption and other security functions, which have been certified by an independent body. However, the EU AI Office acknowledges that some of the protections for whistleblowers under the Whistleblower Directive will only come into effect on August 2 2026.

To the extent the report does not fall within scope of the EU AI Office remit, it will direct the individual to the most relevant authority.

The press release is available here, the FAQs here and the tool here.