Meta has aggressively expanded its superintelligence team by recruiting top AI talent from Google DeepMind and Scale AI, according to LinkedIn data analyzed by Business Insider. Since July, Meta has brought on at least 10 previously unreported researchers from DeepMind, including scientists who worked on Google’s most advanced AI models and even contributed to the company’s gold-medal performance at the International Mathematical Olympiad.
Among Meta’s most significant acquisitions is Yuanzhong Xu, a veteran Google researcher who played instrumental roles in developing LaMDA (Google’s conversational AI model family), PaLM 2 (the large language model foundation for Gemini), and Google’s flagship Gemini generative AI models. Xu now works in what appears to be a specialized unit within Meta Superintelligence Labs dedicated to training the company’s most advanced AI systems.
Meta also hired Tong He, whose research Google credited with achieving gold-medal-level performance at the International Mathematical Olympiad in July. Additional DeepMind recruits include Mingyang Zhang, who specializes in helping large language models retrieve information effectively, and Xinyun Chen, an expert in autonomous code generation and logical reasoning.
Beyond DeepMind, Meta has heavily recruited from Scale AI’s SEAL (Safety, Evaluations, and Alignment Lab) team, bringing on at least six researchers in recent weeks. This follows Meta’s $14 billion acquisition of nearly half of Scale AI in June, which included tapping Scale’s CEO Alexandr Wang to lead Meta’s superintelligence effort.
From Scale’s SEAL division, Meta recruited Ziwen Han and Nathaniel Li, both coauthors of “Humanity’s Last Exam”—a benchmark test designed to be the most challenging assessment humans can create for AI systems. Other Scale AI hires include Alexander Fabbri, Felix Binder, Chen Xing, and Lifeng Jin, many of whom now focus on safety and evaluation issues at Meta Superintelligence Labs.
The SEAL team specializes in crucial AI safety areas, ensuring chatbots align with human values and don’t produce harmful content like misinformation or dangerous instructions. Summer Yue, SEAL’s first leader, joined Meta shortly after the superintelligence team’s unveiling, now leading the alignment group at MSL. Julian Michael, who succeeded Yue at SEAL, also made the transition to Meta.
This recruiting blitz highlights the intense AI talent war among Big Tech companies, with Meta making headlines for eye-popping compensation offers and poaching staff from competitors like OpenAI. Google CEO Sundar Pichai has downplayed the impact, stating retention remains “healthy.”
Key Quotes
We operate one of the most important research teams in AI, and we’re growing it. The moves of a small number of employees to Meta as part of the transaction are a distinct dynamic from the other companies mentioned in this story.
Joe Osborne, a Scale AI spokesperson, provided this statement to Business Insider, attempting to distinguish the employee departures to Meta as part of their business transaction rather than traditional poaching, while emphasizing Scale AI’s continued importance in AI research.
Google CEO Sundar Pichai has previously dismissed the impact of Silicon Valley’s raging talent wars, saying its retention remains ‘healthy.’
Despite losing at least 10 DeepMind researchers to Meta’s superintelligence team, Google’s CEO publicly downplays concerns about talent attrition, suggesting the company maintains confidence in its ability to retain key personnel amid intensifying competition for AI expertise.
Our Take
Meta’s systematic recruitment from DeepMind and Scale AI reveals a sophisticated strategy beyond simple talent poaching—it’s acquiring specific expertise in both frontier model development and AI safety. By hiring the architects of Google’s Gemini foundation alongside safety specialists, Meta is building a vertically integrated superintelligence capability.
The timing is particularly significant. As AI systems approach more powerful capabilities, the combination of cutting-edge research talent and safety expertise becomes essential. Meta’s willingness to spend $14 billion on Scale AI, then immediately absorb its safety team, shows strategic foresight about regulatory and ethical challenges ahead.
This consolidation pattern—where a few tech giants can afford to acquire entire research divisions—may ultimately limit AI innovation diversity. When the best researchers concentrate at Meta, Google, and OpenAI, the industry risks groupthink and reduced exploration of alternative approaches. The brain drain from specialized AI safety startups like Scale AI to Big Tech could particularly concern regulators focused on independent AI oversight.
Why This Matters
Meta’s aggressive talent acquisition from DeepMind and Scale AI represents a critical escalation in the AI arms race among tech giants competing to develop superintelligent systems. By recruiting researchers who built Google’s most advanced models—including Gemini’s foundational technology—Meta is directly challenging Google’s AI dominance while accelerating its own capabilities.
The focus on Scale AI’s safety and evaluation experts signals Meta’s recognition that developing powerful AI isn’t enough; ensuring these systems are safe, aligned with human values, and properly evaluated is equally crucial. This addresses growing concerns about AI safety as models become more capable and potentially dangerous.
The $14 billion Scale AI investment and subsequent talent migration demonstrates how AI development increasingly requires both cutting-edge research capabilities and robust safety infrastructure. Meta’s willingness to pay premium compensation and acquire entire teams shows the existential importance companies place on AI leadership.
For the broader industry, this talent consolidation raises questions about innovation concentration and whether smaller AI labs can compete when tech giants can simply acquire their best researchers. It also highlights the critical shortage of AI expertise, particularly in specialized areas like safety and alignment, making top researchers extraordinarily valuable assets in the race toward artificial general intelligence.
Recommended Reading
For those interested in learning more about artificial intelligence, machine learning, and effective AI communication, here are some excellent resources:
Recommended Reading
Related Stories
- Tech Workers Are the Real Winners in the AI Talent War, With Pay Set to Soar by 2024
- Google’s Gemini: A Potential Game-Changer in the AI Race
- OpenAI’s $157B Valuation: Can It Win the Brutal AI Race?
- Sam Altman Addresses OpenAI Executive Exodus at Italian Tech Week
- The Artificial Intelligence Race: Rivalry Bathing the World in Data
Source: https://www.businessinsider.com/meta-superintelligence-team-hires-deepmind-scale-ai-2025-8