The Vergecast · Friday, September 25, 2026
While Meta's AI agent 'Muse' is acknowledged as being "shockingly useful" and a "good product" for consumer use, deep-seated concerns about Meta's historical data practices and the potential for 'enshittification' remain. Speakers express reluctance to grant the AI access to sensitive information like Gmail and text messages, fearing future data exploitation.
“Um, but it is shockingly useful. And it's, it's a, it's a good product.”
“There's a level of trusting that Meta will eventually enshittify and all of their products will just sort of extract more from you.”
“And that is sort of the problem. So, do you believe I saw an analyst writing, uh, these downloads indicate that Meta's privacy terms and conditions and its promises about secure virtual machines are resonating with consumers. I don't know what you're talking about. Consumers do not know about virtual machines. I guarantee you.”