AI companies race to promise private agents, but doubts remain
At OpenAI DevDay, Sam Altman introduced the Dots agent and said the company aims to set a new privacy standard for frontier AI. OpenAI also criticized Meta’s Muse agent for data safety, even though Meta had marketed Muse as a safer alternative to OpenClaw. The article questions whether AI labs can keep such privacy promises as they encourage users to share more personal information.
At OpenAI DevDay, Sam Altman presented Dots and said it would establish a fresh privacy benchmark for advanced AI. OpenAI used the event to contrast Dots with Meta’s Muse, arguing Muse failed to protect user data. Months earlier, Meta had marketed Muse as a safer successor to OpenClaw, with Mark Zuckerberg emphasizing safeguards and confidentiality.
Muse stores data in an isolated virtual machine, yet Meta can still reach it. A patched zero-day raised outside-control fears, and reported launch issues included possible access to Meta’s internal databases. Muse defaults to training on user inputs unless people opt out; users also described unexpected message reading and address exposure.
If privacy promises don’t hold, users sharing personal data with agents may face surveillance, breaches, or unwanted model training. Competitors could be pressured to match stronger safeguards, while regulators and security researchers may scrutinize claims. Businesses adopting agents may weigh convenience against liability. The outcome could shape trust in AI assistants and determine whether privacy becomes a real differentiator or just marketing.