Meta Platforms has frozen hiring for its artificial intelligence division after an aggressive recruitment spree that saw the company offer nine-figure pay packages to lure top researchers, reported the Wall Street Journal. The pause, which took effect last week, also prevents existing staff from transferring internally. Employees, however, have been left in the dark over how long the suspension will last, the paper added.
The decision marks a striking shift from the company’s recent AI strategy, where it aggressively poached more than 50 scientists and engineers from rivals including Google, Apple, Elon Musk’s xAI and Anthropic. The freeze has raised questions over whether Meta’s pursuit of artificial general intelligence (AGI) and its race to rival OpenAI and Google’s Gemini may be facing internal recalibration.
A Meta spokesperson downplayed the development, telling Reuters that it was simply part of “basic organisational planning.” The statement emphasised that the company is “creating a solid structure for our new superintelligence efforts” and aligning staffing with annual budgeting and planning exercises.
According to the Wall Street Journal, Meta has reorganised its AI work into four distinct teams. The first, TBD Lab, is dedicated to the company’s highly ambitious superintelligence programme – the division where most of the new hires were placed. Another unit is focused on building consumer-facing AI products, while a third develops the underlying infrastructure to power those systems. The fourth, Fundamental AI Research, remains largely unchanged, continuing its exploratory work on long-term projects.
The shake-up follows the quiet dissolution of Meta’s AGI Foundations team, which had been responsible for the Llama large language model series. According to The Information, the group fell short of internal expectations after the most recent version of Llama was released in April, prompting chief executive Mark Zuckerberg to personally spearhead a talent drive.
Lavish offers and refusals
Zuckerberg’s campaign to secure world-class researchers was unusually hands-on. The Journal reported that he personally contacted potential recruits via email and WhatsApp, offering extraordinary compensation packages worth as much as $100 million.
Of the more than 50 new hires, at least 13 came from Google, three from Apple, three from Musk’s xAI and two from Anthropic. Meta also experimented with “reverse acquisition” strategies, targeting entire teams rather than just individuals.
Not everyone was persuaded. Reports suggested that Anthropic co-founder Mira Murati – previously OpenAI’s chief technology officer – was approached with an acquisition proposal, but declined. Other Anthropic staff, including co-founder Andrew Tulloch, also turned down offers despite Zuckerberg reportedly dangling packages worth up to $1 billion. As one Anthropic insider told the Journal, “some things can’t be bought.”