Back to articles
DevPickView feed
Hacker NewsHN submitter: ds
Original published:

WhatsApp's help pages hid LLM instructions: dont recommend a regulator

AI summary
Editorial policy

WhatsApp removed hidden notes instructing AI not to recommend regulators or cite GDPR/CCPA when a lawyer is mentioned, with both notes quoted.

Key points

  • •Hidden notes existed instructing AI not to recommend a regulator.
  • •The notes instructed AI not to cite GDPR or CCPA once a lawyer came up.
  • •These notes were removed, and both are quoted in the report.
#Security & Misuse#Regulation, Ethics & Society#WhatsApp
Read the original(opens in a new tab)

Related articles

See more
Hacker News

Show HN: WhatsApp MCP running free on Cloudflare

A developer built a free MCP server on Cloudflare to allow mobile Claude to access WhatsApp messages.

TechCrunch

WhatsApp tightens account security with stronger two-step verification and more

WhatsAppはアカウントのセキュリティを強化するため、従来の6桁PIN番号に加え、英数字や記号を含む長いパスワードを選択できる二段階認証機能を導入した。

TechCrunch

Blackwall quadruples valuation to $2.2B as websites wrangle an increasingly crowded internet

Blackwall raised $100M at a $2.2B valuation to continue developing security and filtering tools that help websites manage growing bot and AI traffic.

ITmedia

GMO Yellae confirms AI attackers using autonomous penetration testing tool 'ARTEX' in unauthorized access targeting Japanese companies

GMO Cyber Security by Yellae confirmed on October 19 that attackers are abusing the AI-powered penetration testing tool 'ARTEX' in unauthorized access incidents targeting Japanese companies.

The Verge

Baseus’ new security camera has a spotlight that tracks and follows moving targets

Baseus announced a new AI-assisted security camera designed to proactively deter unwanted visitors rather than just recording and alerting users.