Meta CEO Mark Zuckerberg used an internal town hall on Thursday to tell employees that the company's push into artificial intelligence has been moving more slowly than he had hoped.
According to Reuters reporter Katie Paul, Zuckerberg said Meta's development of AI agents — software meant to carry out tasks on a user's behalf — had not accelerated as expected. He also acknowledged that a recent reorganization of the effort was not as "clean" as it could have been.
TechCrunch reported similar remarks, noting that Zuckerberg told staff AI development efforts were not progressing as quickly as anticipated.
Business Insider framed the message around Meta's longer-term ambitions, reporting that Zuckerberg said building "superintelligence" will take time. Business Insider also reported that Meta is making its AI training program opt-in following a data leak.
The comments amount to a candid internal admission from a company that has invested heavily to position itself near the front of the AI race. Rather than touting rapid breakthroughs, Zuckerberg described a slower, bumpier path — both in the specific area of AI agents and in how the work has been organized.
Why it matters: When one of the world's largest and best-funded technology companies acknowledges to its own staff that its AI progress is lagging expectations, it offers a rare, plain-spoken reality check on how difficult and uneven the path to more capable AI systems still is.