Meta Platforms is preparing a fourth shake-up of its artificial-intelligence operations in just six months, The Information reported, citing people familiar with the plans. The company intends to split its year-old Meta Superintelligence Labs into four units covering its Meta AI assistant, core infrastructure, long-term research under the Fundamental AI Research (FAIR) banner and a still-unnamed group focused on the next generation of its Llama large-language model. The reorganisation follows a turbulent period marked by rapid hiring and repeated strategy shifts as Chief Executive Officer Mark Zuckerberg tries to accelerate AI products for the company’s 3.4 billion daily users. Earlier this summer Meta brought in Scale AI founder Alexandr Wang and former GitHub CEO Nat Friedman to co-lead Superintelligence Labs and has poached dozens of researchers from rivals including OpenAI and Google. Investors reacted coolly to the latest overhaul. Meta shares fell about 3% on Monday afternoon in New York, underperforming the broader technology sector as the market weighed whether yet another structural change will translate into faster product delivery and revenue growth.
$META lagging today… https://t.co/pefwN7wAsX
It’s clear GPT-5 has fallen short. Google $GOOGL still holds the lead with the strongest model, and that edge looks likely to continue. xAI’s Grok is also projected to stay ahead of OpenAI, with Grok-5 expected before year-end. https://t.co/vCAw4dNvin
Meta is reportedly planning to divide its Superintelligence Labs segment into four groups: one focused on the Meta AI assistant, an infrastructure team, a long-term research team and a “TBD” team. Some questions remain as Meta goes all in on AI, such as how it can lock in users https://t.co/0gz7xxGeqe