Meta is preparing a major push to regain ground in the artificial intelligence arms race, with plans to invest approximately $15 billion in data-labelling start-up Scale AI and hire its co-founder Alexandr Wang along with key researchers. The deal, which could be announced as soon as Wednesday, would give Meta a 49% stake in the company and value Scale at around $28 billion, according to people familiar with the matter, Financial Times reports.
The investment marks one of the largest deals of its kind in the AI space, underscoring Meta’s urgency to catch up with rivals such as OpenAI, Google, and Anthropic — all of whom have launched increasingly sophisticated AI models in recent months.
The deal is expected to be a cornerstone in Meta’s effort to build a new “superintelligence” lab designed to surpass human intelligence and outperform competitors’ most advanced models. The lab, internally seen as a critical new chapter for Meta’s AI division, aims to overcome the limitations of the company’s underwhelming Llama 4 model, which lagged behind on core reasoning and coding benchmarks.
Neither Meta nor Scale AI has officially commented on the deal. However, insiders told the FT that Meta’s interest is not only in Scale’s data-labelling infrastructure — which is crucial for building high-quality AI systems — but also in recruiting top-tier talent, including Wang himself.
Wang, 28, who co-founded Scale in 2016 and is now a paper billionaire, is expected to play a leading role in the new Meta lab. Although known more for his business acumen and promotional abilities than direct research leadership, he has forged close relationships with AI power players like OpenAI’s Sam Altman and major Silicon Valley investors.
With Wang potentially stepping aside from Scale’s day-to-day management, Jason Droege, currently chief strategy officer and a former Uber Eats executive, is poised to become CEO, sources told the FT.
The implications for Scale’s broader workforce remain uncertain. Wang had previously expressed a desire to take the company public, but this significant investment could delay or alter those IPO ambitions. Internally, Scale has been expanding its focus beyond its original data-labelling niche, increasingly targeting enterprise AI applications and bidding for government contracts in an effort to diversify revenue streams.
Meta’s pursuit of Scale follows recent Big Tech trends: last year, Microsoft paid $650 million to license Inflection AI’s technology and bring in its founder Mustafa Suleyman and team, while Google shelled out $2.7 billion for a similar deal with Character AI. These transactions, structured to avoid triggering regulatory alarms, have nonetheless attracted scrutiny from antitrust regulators — a risk Meta may also face with this move.
