Meta is grappling with significant challenges in its artificial intelligence initiatives, as CEO Mark Zuckerberg acknowledged that progress has fallen short of initial expectations. Despite spending billions on AI development, the company struggles with low morale and has somewhat depended on rival firms’ AI technologies to advance its own tools.
Zuckerberg disclosed during a recent internal town hall meeting that Meta’s efforts to develop AI agents have not accelerated as planned in recent months. He attributed some of these setbacks to timing misjudgments around the company’s restructuring and layoffs, which disrupted operations and failed to deliver smooth results.
The tech giant is investing an unprecedented $145 billion in AI infrastructure this year alone, yet that commitment has yet to translate into the breakthroughs Zuckerberg envisioned. This massive spending highlights the gap between ambitious corporate goals and the technical realities of AI research and deployment.
Alongside development hurdles, Meta recently paused its controversial employee monitoring program intended to collect data for AI training after internal leaks exposed sensitive worker information. The program’s revival will reportedly be based on opt-in participation, according to the company’s CTO, attempting to balance data needs with employee privacy concerns.
Meta’s experience reflects a wider industry struggle: replacing human labor with AI proves more complex and slower than many anticipated. Some companies have even rehired workers they previously let go as AI automation efforts underperform. Zuckerberg remains cautiously optimistic, suggesting that the company might reap AI benefits within a few months, but the current trajectory invites skepticism about whether Meta can regain momentum quickly.