The new initiative aims to unify Meta's AI teams under a single powerful unit led by former Scale AI CEO Alexander Wong.
In a significant move to bolster its position in the competitive artificial intelligence landscape, Meta CEO
Mark Zuckerberg announced the establishment of Meta Superintelligence Labs (MSL), a new and ambitious unit that aims to advance the company's AI research and development.
This new entity will consolidate all of Meta's various AI teams, including those working on foundational models and research through its Fundamental AI Research (FAIR) division.
Alexander Wong, the former CEO of Scale AI, will lead this initiative, supported by a team of prominent talents recruited from leading AI laboratories such as OpenAI and Google.
This restructuring represents a decisive shift in Meta's strategy to compete directly with rivals like OpenAI and Google, as it seeks to accelerate the development of 'superhuman and personalized AI' for billions of users, according to Zuckerberg.
The formation of MSL follows an intensive recruitment drive at Meta, where the company is actively hiring leading experts in the AI field to reduce the gap with its competitors.
In June, Zuckerberg announced the hiring of Alexander Wong as Chief AI Officer, with Wong spearheading the new laboratory.
This initiative is part of a broader strategic investment of $14.3 billion tied to Scale AI, which provides critical data labeling services to many of the industry's leading models.
Additional key talents joining Meta include Nat Friedman, formerly the CEO of GitHub, who will work closely with Wong to focus on AI products and practical research initiatives.
The competition for AI talent has reached unprecedented levels; Sam Altman, the CEO of OpenAI, revealed that Meta attempted to recruit AI researchers from OpenAI, offering signing bonuses of up to $100 million.
Meta's aggressive hiring strategy also includes a lineup of experts from various leading labs, signaling its intent to compete fiercely in the AI sector.
Among those recruited are experts like Tripti Bansal and Shuchao Bi from OpenAI, as well as key contributors from Google's research division.
Zuckerberg noted that Meta is well-positioned to deliver superhuman intelligence globally due to its substantial resources and experience in scaling products that serve billions of users.
The company plans to make significant progress with its foundational model Llama 4.1 and Llama 4.2, which are integral to Meta AI's application, currently benefiting over a billion monthly users.
Meta is also initiating research on the next generation of models, aiming to lead in AI technology advancements within the next year.