Alexandr Wang
Chief AI Officer, Meta
Founding Scale AI
Alexandr Wang grew up in Los Alamos, New Mexico, where his parents were physicists at the national laboratory[2]. He studied at MIT, and told a House committee in April 2025 that he learned there that progress in AI depends on data, compute and algorithms, and that few of his classmates were working on the data problem[2]. He dropped out of MIT to co-found Scale[4].
Scale AI was founded in 2016 and went through Y Combinator's Summer 2016 batch, where Wang is listed as founder and CEO[1]. The company supplies labelled data for training AI models, and set up subsidiary platforms such as Remotasks and Outlier to recruit and manage the gig workers who label it[4]. Y Combinator's company page lists customers including Meta, Microsoft, the U.S. Army, the Defense Innovation Unit, OpenAI, General Motors and Toyota Research Institute, and describes a platform built around reinforcement learning from human feedback[1]. By June 2025 Scale had about 1,500 employees[4].
Positions on AI and national security
In a November 2022 essay titled The AI War and How to Win It, Wang argued that the country that wins in AI this decade will be economically and militarily dominant for the next 50 years[9]. In the same essay he recommended that the U.S. Department of Defense plan to shift 25% of its budget towards AI-enabled capabilities by 2032 and run AI Grand Challenges inside the department[9]. He also wrote that Scale was working with the Department of Defense on products for AI-enabled warfare[9].
On 9 April 2025 Wang testified as Scale's founder and CEO before the House Committee on Energy and Commerce at a hearing on AI, energy and American competitiveness[2]. He set out four themes he called dominate, unleash, innovate and promote[2]. Under them he proposed a national AI data reserve to make government data available for AI work, at least one agentic government programme in each federal agency, a use-case-based regulatory framework with a single federal AI standard, and an expanded role for NIST in exporting AI measurement science[2].
Move to Meta
On 12 June 2025 Scale announced a new investment from Meta that valued the company at over $29 billion, and said Wang would join Meta to work on its AI efforts[3]. Reuters, as reported by RTÉ, said Meta would take a 49% stake for $14.3 billion, according to two sources[4]. Scale's board appointed chief strategy officer Jason Droege as interim CEO, and Wang remained a director on Scale's board[3].
On 30 June 2025 Mark Zuckerberg told staff in a memo that he was creating Meta Superintelligence Labs (MSL) and that Wang had joined as Meta's Chief AI Officer to lead it[5]. Former GitHub CEO Nat Friedman joined to partner with Wang, heading AI products and applied research[5]. In August 2025 Wang announced in an internal memo that MSL would be reorganised into four groups, with a new unit called TBD Lab, run by Wang, focused on foundation models such as the Llama series[6].
Muse Spark and current work
In April 2026 Meta released Muse Spark, the first AI model produced by Meta Superintelligence Labs[7]. Fortune reported that Meta described it as its first reasoning model, that it is multimodal across text and images, and that it powers the Meta AI assistant in the Meta AI app and on meta.ai[7]. Meta presented it as the first in a series of models and said it hopes to open-source future versions, though at launch it was not released as an open-weight model[7]. In March 2026 Meta also created a separate applied AI engineering organisation led by Maher Saba, which works alongside Wang's Superintelligence Labs[7].
Wang writes a Substack newsletter, Rational in the Fullness of Time, which he describes as things he has learned building a company[8]. In a September 2026 post titled Why We're Building Muse, he described the goal as giving every person a second mind that finds out what they want and helps make it happen, including writing plans, sending emails and making calls[8].
Known for
- Founding Scale AI in 2016
- Leading Meta Superintelligence Labs as Meta's Chief AI Officer
- Meta's Muse Spark model, released in April 2026
- Congressional testimony on U.S. AI policy and data
Sources
- Y Combinator: Scale AI company page — checked 6 Nov 2026
- U.S. Government Publishing Office: House Energy and Commerce hearing, Converting Energy into Intelligence, April 2025 — checked 6 Nov 2026
- Scale AI: Founder, Alexandr Wang, Joins Meta to Work on AI Efforts, June 2025 — checked 6 Nov 2026
- RTÉ / Reuters: Meta poaches Scale AI CEO after taking stake in startup, June 2025 — checked 6 Nov 2026
- CNBC: Zuckerberg creating Meta Superintelligence Labs, read the memo, June 2025 — checked 6 Nov 2026
- TechCrunch: Meta is shaking up its AI org again, August 2025 — checked 6 Nov 2026
- Fortune: Meta unveils Muse Spark, April 2026 — checked 6 Nov 2026
- Alexandr Wang, Rational in the Fullness of Time: Why We're Building Muse, September 2026 — checked 6 Nov 2026
- Alexandr Wang, Rational in the Fullness of Time: The AI War and How to Win It, November 2022 — checked 6 Nov 2026