Amazon’s a-Muse-ing beef with Meta
Just two weeks after its launch, Meta’s Muse finds itself at the center of the first big fight over AI agent access.
• 3 min read
TL;DR: Meta’s new personal AI assistant Muse is taking off—and Amazon has already shut it out. It’s a sign that the more AI agents can do, the more pushback they’ll likely get from the platforms they depend on to make them worth your while. And the standoff over who controls the user experience is also raising questions about trust, both for Meta and for agentic AI.
What happened: Meta launched Muse two weeks ago, billing it as a personal agent for people with no technical experience. Muse can handle multistep tasks like sending an email or booking travel, the company said, which means plugging it into your entire digital life (not just Facebook and Instagram). Turns out, people were ready to hand it all over—it became the No. 1 free app on the US App Store last week, app-mogging ChatGPT.
But Amazon wants to be excluded from the narrative: After Meta refused to keep Muse off its site, Amazon blocked the agent for violating its conditions of use, claiming Meta never said Muse would browse the store. The agent doesn’t identify itself, and it appears to get and store customers’ login credentials, Amazon said. (Meta said Muse has “no visibility” into passwords or payment methods and that it can use shared credentials without seeing them.)
Not in my store: Amazon has a lot to lose when an AI agent it doesn’t control does the shopping for you. The company’s ad business, which generated more than $68 billion last year, hinges on shoppers scrolling and seeing sponsored listings, and an outside agent that heads straight to checkout skips all of that.
Tech Brew breaks down the biggest tech news, emerging innovations, workplace tools, and cultural trends so you can understand what's new and why it matters.
By subscribing, you accept our Terms & Privacy Policy.
Trust issues: Handing an AI agent the keys to your digital life means trusting the company behind it, and Meta hasn’t exactly earned the benefit of the doubt. An Oppenheimer survey found that only 8% of US consumers said they’d trust Meta with their passwords, less than a third of the share who’d trust Google. That’s a problem for an agent that connects to your email, your calendar, and Meta’s own platforms (whose algorithms already know what keeps you doomscrolling). It’ll also remember details you’ve mentioned once and suggest actions based on them, per Meta, which could make it a data gold mine.
Bottom line: Meta and Amazon’s standoff is the first step in AI agents’ battle over access. How it plays out could determine what they can actually do for you, given they’re only as useful as the places they’re allowed to go and what they can do there (look no further than this Siri AI blunder). —LC
Also at Meta…
- It’s not a satellite. It’s not a cell tower. Meta’s next big internet bet is a 4,300-mile subsea cable connecting the US and France.
Tech Brew breaks down the biggest tech news, emerging innovations, workplace tools, and cultural trends so you can understand what's new and why it matters.
By subscribing, you accept our Terms & Privacy Policy.