ChatGPT now knows what you do on other websites via ad collector
The highest-engagement AI item of the day tied ChatGPT to website activity collected through advertising, alongside field notes on privacy work and a proxy that scans prompts for secrets.
TL;DR
- The most-engaged AI item of the day, at 708 combined Hacker News points and comments, concerns ChatGPT receiving what users do on other websites through an ad collector.
- Related items include field notes on using LLMs for privacy work and a proxy that scans LLM requests and attachments for secrets before they reach a model.
- The three items point at a common question: model access now carries data-exposure risk that is separate from model capability.
A Hacker News submission titled "ChatGPT now knows what you do on other websites via ad collector" drew 708 combined points and comments, the highest engagement of any AI item in the collection. The claim, as published, is that activity on other sites reaches ChatGPT through an advertising data collector. [1]
Two smaller items approach the same problem defensively. One is a practitioner's field notes on using LLMs for privacy work; the other, preflight, is described as a proxy that scans LLM requests and attachments for secrets before they reach the model. [2] [3]
Why it matters
As assistant features pull in more third-party signals, the privacy boundary of a model session becomes a product decision. The collection shows both the inbound data flow and the tooling being built to inspect it.
Editor's note
This is a claim carried at headline level with no primary documentation in the feed, so confidence is medium; the preflight project and the privacy notes are the concrete, inspectable part of the story.