Meta is establishing a new artificial intelligence research lab aimed at advancing the development of “superintelligence,” a theoretical AI system that would surpass human cognitive abilities.
This move is part of a broader restructuring of Meta’s AI efforts as the company seeks to maintain its competitive edge in the rapidly evolving AI landscape.
The new lab will include Alexandr Wang, the 28-year-old founder and CEO of AI startup Scale AI, who is expected to join Meta following a multi-billion dollar investment deal involving his company. This agreement would also bring additional Scale AI employees to Meta. To bolster its team, Meta has reportedly extended substantial compensation offers to numerous AI researchers from prominent companies like OpenAI and Google, with several accepting the offers.
Under CEO Mark Zuckerberg’s leadership, Meta has committed billions of dollars to becoming a major player in AI. Since the launch of OpenAI’s ChatGPT chatbot in 2022, the tech industry has accelerated investments in AI technology. Meta has integrated AI into various products, including its Meta AI app and Ray-Ban smart glasses.
Meta’s new lab will form part of its ongoing reorganization following recent internal challenges, including management conflicts and staff turnover within its AI teams. Zuckerberg has expressed a strong belief in AI’s transformative potential, describing it as “potentially one of the most important innovations in history” and emphasizing the importance of this year for setting the future course.
The pursuit of “superintelligence” represents a step beyond the goal of achieving artificial general intelligence (AGI)—machines capable of performing any intellectual task that a human can. While companies like OpenAI and Google focus on AGI as a near-term objective, superintelligence remains a longer-term, more ambitious goal.
Meta’s AI research dates back more than a decade, with the company’s first dedicated lab founded in 2013. Yann LeCun, Meta’s chief AI scientist and a prominent figure in neural network research, has led many of these efforts. However, with the recent surge of interest sparked by ChatGPT, Meta has increased its resources, creating new AI teams and launching products like the Llama open-source AI model.
Despite these efforts, Meta has faced challenges, including employee departures and criticism over the performance of its AI models. The release of Llama 4, for example, drew scrutiny after some researchers questioned the validity of Meta’s benchmark comparisons to rival models. This has led Zuckerberg to take a more hands-on approach in recruiting top talent and overseeing AI development.
Zuckerberg’s strategy involves assembling a specialized “superintelligence” team of approximately 50 experts, many personally recruited by him. The new group is expected to work closely with leadership and operate near Zuckerberg at Meta’s headquarters.
Meta’s upcoming investment in Scale AI, a company that specializes in providing data services for AI training, marks one of the largest external investments in its history. Scale AI’s work with major players such as OpenAI and Microsoft positions it as a key asset in Meta’s push for AI leadership.
The formation of the new lab underscores the intense competition among tech giants, including Google, Microsoft, and Amazon, all investing billions into AI innovation and talent acquisition. With AI increasingly integral to the future of technology, Meta is intensifying efforts to develop cutting-edge AI systems that could eventually be integrated across its wide range of products.
Neither Meta nor Scale AI has commented publicly on the new lab or the investment discussions.
The New York Times and Bloomberg contributed to this report.
