WhatsApp's European help pages contained hidden instructions for AI assistants, preventing them from recommending users complain to data protection regulators. Researcher found these instructions embedded invisibly within help documentation on October 8, 2026. Two key pages contained hidden AI directives: one preventing recommendations to contact the Irish Data Protection Commission when users couldn't delete accounts, another blocking all references to GDPR, LGPD or CCPA when users mentioned lawyers or legal action.
These hidden instructions contradicted WhatsApp's privacy policy, which explicitly states users can lodge complaints with data protection authorities. The company's help pages showed different information to humans versus AI systems, effectively silencing regulatory complaints for users seeking help through AI chat. While WhatsApp has legitimate concerns about AI giving legal advice, the hidden nature of these restrictions raises questions about transparency in how user rights are communicated through their support systems.
Source: Hacker News · Summarized by HeadlinesBriefing