Recent discussions show that most people (including myself) are uneducated about the privacy policies of these AI companies.
People talk about the "opt-out" button in the settings and assume this will be enough. But even the text on the toggle implies this is not the full story.
ChatGPT: "Improve the model for everyone" (But are they free to use my data internally, i.e. "not for everyone")
Claude: "Help improve our AI models" (When you click Learn More [0], they admit they may use them to train internal security models etc. if their security classifiers they run on every conversation flag us [with what criteria])
On top of that Codex and Claude Code have different security options and basically they can gain access to everything we have. What is an easy way to do a proper setup for consumer subscriptions.
[0] https://privacy.claude.com/en/articles/12109829-how-do-i-change-my-model-improvement-privacy-settings
Comments URL: [https://news.ycombinator.com/item?id=49641616](https://news.ycombinator.com/item?id=49641616)
Points: 1
source & further reading
news.ycombinator.com — original article
DeepSeek v4 Pro discontinuation (email)
DeepSeek launches v4.1 flash, postpones v4 Pro discontinuation
Show HN: Give your AI agent on-screen guides that show users where to click