AI Daily Grind // AI news, freshly ground daily
Cover illustration for Saturday, October 24, 2026
Morning Token 03 ⚖️ Policy

Big Tech's fine print: your Gmail for ads, your fault for Meta's AI

Google updated its privacy policy. It used to promise not to use your Drive, Gmail or Google Photos for personalized ads. Now there is an exception if you turn on its personal AI features, called Personal Intelligence. Android Authority spotted the change first.

Meta is also changing its rules for businesses. After pushback from news publishers worried about AI scraping, Meta delayed its new business terms to November 20. This is new since yesterday. Users on Reddit also point out that the terms make the customer legally responsible if Meta's AI makes a mistake.

So the pattern is clear. AI features come with new data deals and new legal risks, often hidden in long policy pages.

Why it matters

The more helpful the AI, the more personal data it needs, and that data can turn into ad targeting. Read the settings before you click "turn on".

  1. 01 OpenAI cuts its frontier models off from the live internet Safety
  2. 02 Kimi K3.1 Pro: near the top, half the price, weights promised Launch
  3. 04 Hacking gets cheaper: $50 per website and a hacked iPhone Safety
  4. 05 Block's Dessa: an AI trained only on human conversation Research
  5. 06 Take back your smart TV with open source Open source
  6. 07 Bubble talk grows as Anthropic's IPO gets close Business