Inside Meta: Why Mark Zuckerberg and Alexandr Wang are clashing over the company’s AI future

Meta

• Zuckerberg and Wang clash over Meta's AI strategy and leadership.
• Wang's team clashes with Meta veterans over AI priorities, deepening divides.
• Zuckerberg launches Meta Compute, tightening AI infrastructure control.

Did our AI summary help?

Meta’s big bet on artificial intelligence is starting to show visible cracks at the top. A new internal reporting structure has brought to light growing tension between CEO Mark Zuckerberg and his most expensive hire so far, Alexandr Wang, the 28-year-old founder recruited in a deal valued at around $14 billion.

According to people familiar with the situation, Wang is increasingly frustrated with Zuckerberg’s management style. While the Meta CEO is known for being deeply hands-on, Wang reportedly finds the constant involvement stifling. At the same time, several employees inside Meta are quietly questioning whether Wang has enough experience to lead what is effectively the company’s $600 billion AI ambition.

The friction became hard to ignore during a key strategy meeting this fall. Wang’s TBD Lab team clashed with long-time Meta leaders Chris Cox and Andrew Bosworth. The disagreement was fundamental. Cox and Bosworth wanted Meta’s new AI models to be trained using Instagram and Facebook data to quickly improve ads and core products. Wang strongly disagreed. His view was that Meta first needs to build world-class AI models that can rival OpenAI and Google, even if that means delaying direct product integration.

That difference in priorities has deepened an “us versus them” culture inside the company. Wang’s researchers see themselves as working on the future of superintelligence, while Meta’s old guard is viewed as too focused on feeds, ads, and short-term results. Rumours have also circulated internally about large sums of money being redirected to Wang’s team, though Meta has denied specific figures.

Leadership doubts grew sharper when Yann LeCun, Meta’s chief AI scientist and a Turing Award winner, left in November rather than report to Wang. Sources say LeCun was uncomfortable taking direction from someone whose background was in data labelling rather than core AI research. Around the same time, teams working on products like Vibes, Meta’s AI video feed, felt pressured to move too fast to match competitors like OpenAI’s Sora.

Ironically, this tension may have been inevitable. Zuckerberg hired Wang precisely because he wanted someone young and ambitious to execute his vision of “personal superintelligence.” But Zuckerberg’s intense focus on AI has started to look less like support and more like control.

That became clearer this week when Zuckerberg announced Meta Compute, a new top-level initiative reporting directly to him. By placing key infrastructure decisions above Wang, the move signals that Zuckerberg is tightening his grip as Meta spends billions on AI. For now, it’s clear that Meta is all-in on AI—but exactly who gets to steer that bet remains an open question.