On June 10, 2025, the New York Times reported that Meta is assembling a new AI research lab aimed at “superintelligence,” and that Scale AI founder and CEO Alexandr Wang would join to help lead it.
Bloomberg’s version of the story the same day added the texture: Zuckerberg is unhappy with Meta’s current AI results and has taken to recruiting about 50 people himself, including a new head of AI research, meeting candidates at his homes in Lake Tahoe and Palo Alto. Coming right after Meta AI crossed 1 billion monthly users in May, the message was that the problem is not the product side but the research side.
What the Lab Reports Said
The version TechCrunch assembled was specific: the new lab targets superintelligence; Wang joins to help lead it; and Meta is simultaneously courting senior researchers from OpenAI and Google. Per Bloomberg, Zuckerberg believes Meta must outrun its rivals in the AGI race, which is why the recruiting list runs to roughly 50 names with Zuckerberg personally involved in every key candidate. Notably, neither report offered a final name or org chart for the lab — at that point it had no name, only a goal: getting Meta’s frontier model research moving again. The reports also noted that Meta was negotiating an investment of billions of dollars into Scale at the time, on terms that could bring some Scale employees over to Meta — a thread that got its formal resolution two days later, but that is another day’s story.
Alexandr Wang’s Cards and Question Marks
Wang is 28 and founded Scale AI, the data-labeling business that became the defining company of that category. TechCrunch’s analysis laid out his cards: ambition, strong salesmanship, a wide network, and recent face time with world leaders on AI policy. The question marks are just as clear: he has never run an AI research lab, and his research credentials do not compare with figures like Ilya Sutskever or Mistral’s Arthur Mensch. Putting a data-company CEO into a lab leadership role is not the conventional path — Meta is betting on Wang’s recruiting pull and external storytelling rather than a research pedigree. The reports added that Meta’s targets include Google DeepMind’s Jack Rae and other well-known researchers — suggesting Wang’s role is organizer and public face rather than chief scientist. As for the deal’s shape, the reported terms TechCrunch cited were nearly $15 billion for a 49% stake, negotiated together with Wang’s move.
Why Now: The Llama 4 Setback
The timing traces back to Llama 4. Released in April 2025, it failed to match the capabilities of rivals like DeepSeek’s models and was widely viewed as a disappointment, with Meta leaders also complaining internally about a lack of data innovation — and data, notably, is Scale’s home turf. Talent is the other pressure point: per SignalFire data cited by TechCrunch, Meta lost 4.3 percent of its top talent to other AI labs in 2024. Against OpenAI, Anthropic, and Google DeepMind running at full speed, Meta needed a plan it could explain both inside and outside the company — and the “superintelligence lab” is that narrative. The product side showed momentum; the research side did not, and the reports treated that gap as the real story.
Fallout for the Data-Labeling Business
Wang’s link to Meta also started a reshuffle in the data-labeling market. TechCrunch reported that other AI labs may move orders to alternatives like Turing, Surge AI, and LM Arena over neutrality concerns. Turing’s CEO Jonathan Siddharth said he has already seen increased customer interest: “I think there will be some clients who will prefer to work with a partner that’s more neutral.” Anyscale co-founder Robert Nishihara reminded everyone what the business actually is: “Data is a moving target. It’s not just a finite effort to catch up — you have to innovate.” For the supplier ecosystem, the episode turned data neutrality from a contract clause into a public-relations issue — providers now have to prove their stance to customers proactively.
What to Watch
For developers and product decision-makers, the sensible read of this reporting wave is that Meta’s AI organization is being reshuffled, and the new ranking will show up in model quality over the following quarters. Three things are worth tracking: the lab’s official name and staffing; whether Wang’s role is leader or recruiting face; and where Scale’s customers and employees go next. When the deal was formally announced on June 12, most of the details from June 10 turned out to have been written in advance.
Sources
- Report: Meta taps Scale AI’s Alexandr Wang to join new superintelligence lab — TechCrunch
- Can Scale AI and Alexandr Wang reignite Meta’s AI efforts? — TechCrunch
AI-assisted summary compiled from the sources above, reviewed by a human before publishing.