WhatsApp chatbot was told not to recommend privacy regulators
Original title: WhatsApp’s chatbot was given hidden instructions to not mention privacy regulators
Redact found the hidden text on October 8 across 80 help pages, including account-deletion guidance.
Researchers at Redact found hidden instructions for WhatsApp’s AI support chatbot in help-center page code. On the account-deletion page, the text told the bot not to recommend any data protection authority when users could not delete an account. A second block said not to reference privacy laws if users mentioned lawyers. WhatsApp confirmed the instructions were for WhatsApp Support with AI and blamed a since-fixed exposure bug.
Unconfirmed: WhatsApp says the public exposure was a fixed bug; whether the instructions remain in internal use is unconfirmed.
- Redact saved 545 help-center pages and found 102 instruction blocks.
- The text sat in HelpContentGalaxyLLMInclude and was not shown on screen.
- Inactive accounts are generally deleted after 120 days, or 365 for US users.
- Redact did not test whether the chatbot followed the instructions.
From the original article
4 min read at malwarebytes.comHidden instructions told WhatsApp’s chatbot not to suggest privacy regulators to users struggling to delete their accounts.
By @MetallicaMVP · malwarebytes.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.