AI Daily Grind // AI news, freshly ground daily
Cover illustration for Friday, November 6, 2026
Morning Token 01 🛡️ Safety

Claude Code reportedly sends your email address to websites

Claude Code is Anthropic's AI coding assistant. It runs on your computer and can visit websites for you. A bug report on Anthropic's own GitHub page has a short and worrying title: "Claude Code sends your email address to websites." It was the top AI story on Hacker News, with 1,140 points and 346 comments.

Our sources only give the title of the report. We do not know yet how often it happens, which websites get the address, or why. We also have no reply from Anthropic in our sources. So treat it as a user report, not a confirmed leak.

It is the second Claude Code security story in a week, after the backdoored extension on Cursor last Saturday. If you use coding agents, check what they send out. A new free tool shared on Hacker News, Stroq, makes a local audit of what your coding agent did and why.

Why it matters

Your email address is how spam, phishing and tracking find you. Coding agents run with a lot of access, and small leaks add up fast. This is exactly the kind of problem that needs a quick, public answer, not one months later.

More on: 🛡️ SafetyPrivacyAI agents

  1. 02 OpenAI: GPT-6 Astra quietly does worse on safety research Safety
  2. 03 GPT-6 Cyber stays in bounds more often, and agent security gets pricey Safety
  3. 04 After the midterms: tech stocks fall, a call for a pause, and no more secret NDAs Policy
  4. 05 OpenAI eyes a 2027 IPO, and doubts about Anthropic's $1.5 trillion Business
  5. 06 Agents move into your bank, your job and your trips Launch
  6. 07 Deepfake scam ads, and a patent to change live words Safety