Meta's Muse AI assistant works, but its autonomous data gathering creeps users out
The Verge AI · rss · 2026-09-10
Meta launched Muse, its first real AI productivity assistant, aiming to "take the busywork off your plate" — shopping, emails, trip planning. The Verge's hands-on review found it generally worked as expected, but the unnerving amount of information the agent autonomously gleaned about the user overshadowed the experience. It marks a pivot for a company that previously prioritized entertainment over productivity, with privacy boundaries still untested.
More from Safety
- MacAskill compiles years of extinction warnings from Altman, Sutskever, and Anthropic leaders — kipperrii · 2026-09-10
- Safety researcher: robust AI verification matters more than treaties for global safety — davidmanheim · 2026-09-10
- Open source won't save cybersecurity: whoever has more GPUs wins — andrew_n_carr · 2026-09-10
- From Idiot Savant to Runaway Swarm: How We Got Here and What's Next for AI Risk — stritefax · 2026-09-10
- Podcast: China's agent rules and Baseten's latest — The Cognitive Revolution · 2026-09-10
- Europol lobbies EU to ban end-to-end encryption, labels PETs 'tools of criminals' — wen_ragnarok · 2026-09-10