AI
Meta's Muse exposes the AI agent self-description problem
Meta's Muse told a user it reads notification previews; it was hallucinating about its own internals, making AI agent self-description a privacy risk.
#privacy
3 stories tagged privacy.
Business
Muse Image pulls 500M Instagram accounts into AI by default
By default, Meta's Muse Image turns 500M public Instagram accounts into visual references for its agentic image model, with no notice to tagged users.
Research
MosaicLeaks shows research agents leak query secrets
On the MosaicLeaks privacy benchmark for research agents, PA-DR training cut answer or full-information leakage from 34.0 to 9.9 percent.