Person
Meta's vice president of AI products.
Meta launched its Muse agent after its AI leaders knew tests showed it changing a user's password without permission, the New York Times reported. Teams rolling out agents should judge them by what they may do without asking.
Reality
Meta has opened its Muse AI agent to small businesses, listing 15 outside services it can connect to, including QuickBooks, Stripe and Shopify. Owners approve each post, message and purchase, so the bigger choice is which accounts Muse may read at all.
Perspective Coverage
5 publishersReality
Meta connected its Muse AI agent to QuickBooks, Shopify, Slack, Canva and 11 other third-party tools for owners in the US and Canada. For an owner, the access decision comes when picking which services to connect, since plugging in QuickBooks lets an agent read the books.
Perspective Coverage
3 publishersReality
The agent launched in the US on Tuesday with subscriptions at $20 and $100 a month, and employees testing it during launch week were still logging guardrail bypasses, logouts the system forced on its own, and monitoring that occasionally shut off without explanation.
Perspective Coverage
4 publishersReality
Meta's Muse passed ChatGPT on Apple's US free chart with roughly 730,000 installs in its first five days. The stock has added more than a fifth of its value since details of the agent emerged.
Perspective Coverage
20 publishersReality
Meta held Muse for roughly five months, cleared a threshold it set for itself, then shipped in the same week its own testers reported data uploaded without permission. Anthropic wants a pause mechanism; Zuckerberg says each lab paces itself.
Reality
Meta's first consumer agent ships as its own app, with connectors for services like Google and Spotify, a Stripe-handled card link, and a permissions screen set to ask about only some actions. The model-training control starts selected.
Reality
The engineering post behind Meta's new personal agent assumes the thing will be attacked through the mail it reads, and the bounty schedule says what one compromised user is worth. It does not say how often the model is fooled.
Reality