Meta's Muse Auto-Opts Users Into AI Data Collection — SkimNews
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Meta's Muse app surpassed 900,000 downloads in its first week, per Sensor Tower, and operates as a personal AI agent that browses the web via a "virtual machine" to order food, find Marketplace couches, and message sellers
- Muse users are automatically opted in to having their interactions used for AI model training, with Meta saying the data is "sanitized" but offering no toggle to disable the app's memory feature entirely
- Meta's own safety blog acknowledges Muse's agent actions—like booking a dinner reservation or buying a Marketplace item—could "indirectly influence" the ads users see on Instagram
- Consumer advocates flagged the data-collection push as a major red flag: Calli Schroeder of EPIC compared it to Meta's history of "manipulative tactics," and Rory Mir of the EFF warned that "when you talk to an AI, you are talking to the company hosting the AI"
- Researchers including Hugging Face's Margaret Mitchell warn the design "disengages users" and risks "cognitive degradation," since agents with access to email, finances, and calendars outsource decisions users would otherwise make themselves
- Meta plans to release "confidential" versions of its virtual machine later this year that "cryptographically and verifiably prevent" the company from accessing user data, while Chief AI officer Alexandr Wang hinted at non-ad revenue opportunities for Muse without disclosing specifics
Why it matters: Meta's automatic opt-in to AI training, paired with aggressive prompts to link financial and personal data, adds a new privacy liability to its track record—with 900,000 users onboarded in week one, the gap between current practices and the "confidential" virtual machines slated for later this year creates an exposure window that consumer advocates say mirrors past broken-trust episodes like the recent Instagram AI deepfake rollout.
Ask SkimNews
