● Covered by 2 sources · 2 reportsMedium impact2 negative
WhatsApp help pages contained hidden AI instructions to avoid mentioning regulators or laws
Key points
- WhatsApp help pages contained hidden AI instructions.
- AI was told not to recommend regulators or mention laws.
- Instructions activated if a user mentioned a lawyer.
- Contradicted WhatsApp's public privacy policy.
- Instructions were removed by October 18.
- Redact, a data-deletion software company, found the text.
- WhatsApp told WIRED instructions were for chatbot reached by messaging WhatsApp Support with AI.
- WhatsApp stated the instructions appeared due to a bug, now fixed.
- A second hidden text block told chatbot not to cite privacy laws if user mentioned lawyer, lawsuit, or court.
- A WhatsApp account can only be deleted from inside the app on the user's main phone.
Hidden AI Instructions Discovered
Hidden instructions for an AI were found within WhatsApp's help pages. These directives specifically told the AI not to recommend data protection regulators or cite relevant laws, particularly if a user mentioned a lawyer. The instructions were embedded in a way that was not visible to users but accessible to an AI system.
Specific Directives to AI
One set of instructions prevented the AI from suggesting a regulator if a user could not delete their account, especially if they had lost their phone or number. Another directive told the AI to avoid discussing legal matters, including privacy laws, once terms like 'lawyer' or 'legal action' were used by the user. This applied even if the user initiated the discussion about the law.
Contradiction with Public Policy
These hidden AI instructions directly conflicted with WhatsApp's publicly available privacy policy for the European region, effective July 2, 2026. The public policy explicitly informs users of their right to complain to a data protection regulator and references the laws that grant these rights, such as Article 17 and Article 77 of the EU regulations.
Scope of Hidden Text
Analysis of 545 WhatsApp help pages revealed 80 pages containing hidden AI text, comprising 102 blocks in total. While most blocks were deemed harmless, the two specific sets of instructions regarding regulators and legal information stood out due to their implications for user rights and access to information.
Updates
- Instructions were removed by October 18.
- Redact, a data-deletion software company, found the text.
- WhatsApp told WIRED instructions were for chatbot reached by messaging WhatsApp Support with AI.
- WhatsApp stated the instructions appeared due to a bug, now fixed.
- A second hidden text block told chatbot not to cite privacy laws if user mentioned lawyer, lawsuit, or court.
- A WhatsApp account can only be deleted from inside the app on the user's main phone.
✨ This summary was generated by AI from the outlets' reporting listed below. It is not independently verified and may contain errors — check the original sources. How BrevFeed works →
The daily brief
One email each morning: the day's tech stories, clustered across outlets and summarized. No account needed.
One email a day. Unsubscribe in one click, any time.
Today's brief
Spend a few minutes, get the whole day. Every topic's top stories in one hands-free rundown — listen, watch, or read the transcript.
▶ Play today's brief
New every morning, and the back catalogue is archived by date.
How outlets covered it
WhatsApp's help pages contained hidden instructions for its AI support chatbot, telling it not to recommend data protection authorities to users who could not delete their accounts. These instructions contradicted WhatsApp's European privacy policy, which states users can complain to such authorities.
WhatsApp's help pages included hidden instructions for an AI, telling it not to recommend data protection regulators or cite relevant laws if a user mentioned a lawyer. These instructions contradicted WhatsApp's public privacy policy for Europe, which outlines users' rights to complain to regulators.