Yhn TechnologyAI first seen 8 d ago, last 1 d ago, peak #1
WhatsApp help pages hid instructions telling AI not to name regulators
Original: WhatsApp's help pages hid LLM instructions: dont recommend a regulator
WhatsApp's help pages contain hidden instructions apparently aimed at large language models, including a line telling the AI not to recommend contacting a regulator. The text, which surfaced on a developer blog, suggests companies are quietly embedding prompts into public documentation to shape how AI assistants summarise or recommend their services, raising questions about transparency and how users are steered away from complaints channels.
Why now: Hidden prompt instructions steering AI away from suggesting regulators raise fresh concerns about corporate manipulation of AI tools.
Rank over time, top of the chart is #1. 47 snapshots from 2 d ago to 1 d ago.
Evidence
API: https://socialmediatrends-api.osmike.com/v1/trends/2930799