Meta CEO Mark Zuckerberg admitted at a general meeting this week that the company’s massive restructuring had not delivered the expected results. According to a Reuters report, Zuckerberg said the company’s AI agents were not developing as quickly as expected and that senior management “miscalculated” the timing of the changes. The relaunch, announced earlier this year, included laying off about 8,000 employees—roughly 10% of Meta’s global workforce—and moving 7,000 employees to AI-focused teams. The moves were aimed at funding Meta’s massive investment in artificial intelligence infrastructure, projected to reach $145 billion in 2026, but have sparked employee resistance and concerns about morale.
Slower Than Expected AI Progress
Zuckerberg also acknowledged that “the agent trajectory over the past four months has not accelerated as much as we expected.” He added that while executives were “over-optimistic” about tools like Anthropic’s Claude Code earlier this year, the company’s bet on AI agents has yet to pan out. Despite the setbacks, Zuckerberg told employees he expects Meta to start seeing greater benefits from its investments in artificial intelligence within three to six months.
Mouse Tracking Software Review
Meta CTO Andrew Bosworth also raised concerns about the company’s controversial mouse tracking software, which was suspended following reports of sensitive data being exposed. Bosworth said the review found that no employee data was included in artificial intelligence training.If reinstated, the program would be opt-in, a reversal from April, when U.S. employees were told there was no option to opt out. Bosworth said: “For people who are comfortable with it, it’s great they can contribute to such a large human research effort. For people who aren’t comfortable with it, it’s not a problem.”
Meta CTO Andrew Bosworth admitted the company had done a “terrible” job rolling out its new applied artificial intelligence division.
Meta’s CTO Andrew Bosworth recently admitted in a memo to employees that the company had done a “terrible” job rolling out its new applied artificial intelligence division, Wired reported. The division, formed in March this year with 6,500 engineers and product managers, was designed to accelerate Meta’s generative AI projects. Instead, workers are calling the transition chaotic, with some calling it a “gulag.” Bosworth added that the company had undermined trust by failing to clearly communicate its vision, destabilizing teams through rapid restructuring. The report also added that Bosworth promised to revive a more cheerful internal culture through improved communication, career opportunities and even office perks. Managers will be limited to 20 direct reports, employees will have access to “AI coaching” tools, and Meta will invest in micro-kitchens, travel budgets and social events to boost morale. “I hope we can recapture the best of the culture we have joined,” Bosworth wrote.