Inside Meta's Grand Strategy to Win the Global AI Race

Meta Platforms this week hired two senior AI researchers, Mark Lee and Tom Gunter, from Apple’s AI division to join its newly formed Superintelligence Labs team. Lee has already commenced work on Meta’s Menlo Park campus, while Gunter is slated to start shortly, Bloomberg reported.

The hires come days after Apple's head of AI models, Ruoming Pang and illustrate Meta’s willingness to beat AI rivals with “mega‑offers” that include multi‑million‑dollar signing bonuses and equity grants.

Meta Superintelligence: The Dream Team

Meta’s hiring spree extends well beyond Apple. It builds on a broader campaign that began in June when Meta invested approximately $14.3 billion for a 49% stake in data‑labeling startup Scale AI, bringing its founder and CEO, Alexandr Wang, onboard as Meta’s chief AI officer.

CEO Mark Zuckerberg’s Meta Superintelligence Labs is designed to bridge the gap between fundamental research, product development and applied AI. In an internal memo, Zuckerberg directed that the new lab report directly to him, with Wang and Friedman spearheading efforts to forge “artificial general intelligence” capabilities that could eventually surpass human reasoning.

The group is tasked with developing advanced language models, autonomous agents, and AI‑driven applications embedded across Meta’s family of apps, from Instagram to Horizon Worlds.

Meta’s Poaching Attempts

Under Wang’s leadership, and alongside former GitHub CEO Nat Friedman, who will co‑lead the unit, Meta has poached researchers from top labs including OpenAI, Google DeepMind, Anthropic, and Apple. Key recruits include Daniel Gross, ex‑Safe Superintelligence CEO; Ruoming Pang, former head of Apple’s Foundation Models team; and several architects of large‑scale language models such as Huiwen Chang and Ji Lin.

The social media giant’s hiring spree shows Zuckerberg’s efforts to close the gap with AI frontrunners after its flagship Llama 4 model fell short of expectations.

By committing “hundreds of billions of dollars” to build multi‑gigawatt data centers and targeting more than 1 million GPUs by 2027, Zuckerberg aims to pair unmatched compute power with world‑class human capital.

Investors and analysts note that such scale is essential for training next‑generation AI systems, while the lavish compensation packages serve both to attract talent and signal Meta’s seriousness in the global AI arms race.

Global AI Race

Meta’s talent hunt is mirrored by other tech giants eager to boost their AI capabilities. Alphabet’s Google recently completed a $2.4 billion deal to license code‑generation technology from startup Windsurf, simultaneously hiring CEO Varun Mohan and core R&D staff to strengthen its Gemini project.

These “exploding offers” and strategic acquisitions reflect a broader shift away from traditional M&A toward selective equity stakes and targeted poaching, as companies seek rapid capability gains while sidestepping regulatory scrutiny.

As the US, China, Europe and other regions pour resources into AI, the struggle for talent has become as critical as breakthroughs in model architectures or chip innovations.

In a recent podcast, Priscilla Chan highlighted that access to powerful GPUs has become a key recruitment lever, an argument echoed by Zuckerberg as Meta scales its fleet to over 1.3 million units by year‑end. Industry projections estimate cumulative AI infrastructure spending could top $1.7 trillion by 2035, making the control of both compute and human expertise central to future technological and economic leadership.

Meta’s bold play for top AI talent, and its corresponding infrastructure investments, highlights a pivotal moment in technology. Whether these efforts translate into commercial products or broader societal benefits remains to be seen, but one thing is certain: the next breakthroughs in AI will be driven as much by where the world’s brightest minds choose to work as by the silicon they run on.

As competition intensifies, the race to recruit, poach, and acquire talent may well define which companies, and which nations, emerge as leaders in the coming era of superintelligent systems.