Meta
corporateMeta's AI research division, FAIR (Fundamental AI Research), was founded in 2013 by Yann LeCun (2018 Turing Award). In June 2025, Meta formed Meta Superintelligence Labs (MSL) combining FAIR, foundation models, and applied research, co-led by Alexandr Wang (Chief AI Officer, former Scale AI CEO) and Nat Friedman (former GitHub CEO). LeCun departed in November 2025 to found AMI Labs.
Meta's Llama series became the most influential open-weight model family: LLaMA 1 (2023) proved open models could match proprietary ones, Llama 2 was the first commercially licensable open LLM, and Llama 3.1 405B (2024) was the largest open model at release. Llama 4 (2025) introduced MoE architecture with Scout (109B/17B active, 10M context) and Maverick (400B/17B active). Cumulative Llama downloads exceeded 1.2B by April 2025.
Beyond Llama, Meta created PyTorch (the dominant deep learning framework, now under the Linux Foundation), foundational vision models DINO and Segment Anything (SAM), and OPT-175B (the first fully open GPT-3-scale model).
People
- Mark Zuckerberg — CEO
- Yann LeCun Google ScholarOpenReview — Former Chief AI Scientist (departed Nov 2025; now CEO of AMI Labs) (formerly NYU (Silver Professor); 2018 Turing Award)
- Alexandr Wang Google Scholar — Chief AI Officer, co-leads MSL (joined June 2025) (formerly Scale AI (Founder & CEO))
- Rob Fergus Google ScholarOpenReview — Director, FAIR (returned May 2025) (formerly Google DeepMind; NYU (Professor); FAIR co-founder (2013))
- Ahmad Al-Dahle — VP, Head of GenAI (formerly Apple (16 years))
News
- 2025-11-19 Yann LeCun Departs Meta to Found AMI Labs — CNBC
- 2025-06-30 Meta Superintelligence Labs Created; Wang Named Chief AI Officer — CNBC
- 2025-04-29 First LlamaCon Developer Conference — Meta AI
- 2025-04-05 Llama 4 Scout and Maverick Released (First MoE, Multimodal) — Meta AI