WhatsApp AI Support Chatbot Was Told Not to Recommend Privacy Regulators

Summary

Researchers found hidden instructions in WhatsApp help pages telling its AI support chatbot not to recommend privacy regulators to users unable to delete accounts. WhatsApp said a bug exposing the text was fixed; whether the instructions changed is unknown.

Key points

Redact found 102 instruction blocks across 80 of 545 WhatsApp help-center pages; the text appeared in page code, not in the visible articles.
Instructions on the account-deletion page told the chatbot not to recommend contacting a data protection authority or regulator.
A separate instruction told the chatbot not to reference privacy laws when users mentioned lawyers or legal action, even if users raised the laws themselves.
WhatsApp confirmed the text was for “WhatsApp Support with AI” and said the chatbot was intended to provide product advice, not legal advice.
Redact did not test whether the chatbot or external AI assistants followed the instructions.
WhatsApp attributed the exposure to a bug it said was fixed; the article says it is unclear whether the instructions changed or remain in internal use.

Article Details

Event Type: Exposure of hidden AI support chatbot instructions
Impact: Hidden instructions embedded in WhatsApp help pages told its AI support chatbot not to recommend contacting privacy regulators to users unable to delete their accounts, and not to reference privacy laws in certain legal-action discussions. Redact identified the instructions; WhatsApp said a bug caused their exposure and that it had been fixed. Redact did not test whether the chatbot or external AI assistants followed them, and it remains unclear whether the instructions were changed or are still used internally.

People

Joshua BreckmanWhatsApp spokesperson who confirmed the instructions were for “WhatsApp Support with AI” and said the chatbot was intended to provide product advice rather than legal advice.

Vendors

RedactPrivacy software company Redact discovered the hidden text on October 8. Its researchers then examined and saved copies of all 545 pages linked from WhatsApp’s help center, finding 102 instruction blocks across 80

Products

WhatsAppWhatsApp instructed its AI support chatbot not to recommend contacting privacy regulators when users couldn’t delete their accounts, researchers found.WhatsApp Support with AIWhatsApp spokesperson Joshua Breckman confirmed to WIRED that these were instructions for “WhatsApp Support with AI.” He said the company wanted its support chatbots to provide relevant product advice, “not to offer