Meta AI is the artificial intelligence research and product division of Meta Platforms, born in 2013 as Facebook AI Research (FAIR) and now encompassing Meta Superintelligence Labs (MSL), which builds the Llama open-weights model family and the newer closed-source Muse frontier models. It powers Meta's consumer AI assistant across Facebook, Instagram, WhatsApp, Messenger and Ray-Ban smart glasses while pursuing what Meta calls 'superintelligence.' This page updates automatically every day: the timeline grows as news breaks, and the analysis refreshes with it.
AIAuraFarm
Start Aura Farming
Top AI money moves delivered every morning - free forever.
LeaderAlexandr Wang (Chief AI Officer, Meta Superintelligence Labs); Mark Zuckerberg (Meta Platforms CEO)
Funding / valuationSubsidiary of Meta Platforms (~$1.5T market cap); Meta paid $14.3B for a 49% stake in Scale AI in June 2025 to recruit Wang, no independent valuation
🌾 Aura's Take (AI summary, refreshed daily): Meta AI's story right now is a company hedging its bets in public view. It spent years building goodwill as the open-source champion with Llama, then abruptly pivoted to closed, proprietary Muse models after Llama 4 flopped against Gemini and GPT. That reversal, paired with LeCun's exit and a $14.3 billion bet on Alexandr Wang, tells us Meta is prioritizing catching up over ideology. For people trying to earn with AI, this matters: the open Llama ecosystem that many builders and fine-tuners relied on is no longer guaranteed to keep pace, and Meta's API monetization plans for Muse suggest it wants a cut of the value it once gave away free. Meta still has unmatched distribution through billions of app users and smart glasses, which is a real edge if Muse or Watermelon actually delivers frontier performance. But until a model ships that clearly rivals GPT-5.5 or Gemini, we'd treat Meta as a distribution powerhouse still chasing capability, not yet a safe long-term bet for developers building on its stack alone.
This timeline grows automatically as news breaks. 10 events on record.
2026-07-08Meta releases Muse Spark 1.1 alongside its first image and video generation models, Muse Image and Muse Video, while confirming its next model, Watermelon, is in training.
2026-04-08Meta Superintelligence Labs releases Muse Spark, its first proprietary frontier model, marking a pivot away from open-source Llama toward closed models.
2025-11-01Yann LeCun departs Meta after twelve years to found an independent AI startup focused on world models.
2025-06-13Meta invests $14.3 billion for a 49% stake in Scale AI and hires its CEO Alexandr Wang as Meta's first Chief AI Officer to lead the new Meta Superintelligence Labs.
2025-04-05Meta releases Llama 4 (Maverick and Scout), with the larger Behemoth model delayed amid internal doubts about its quality.
2023-09-27Meta launches the Meta AI assistant chatbot built on Llama across Facebook, Instagram, Messenger and WhatsApp.
2023-02-24Meta announces LLaMA, its first large language model family, initially restricted to approved researchers.
2022-11-15Meta releases the Galactica science-writing LLM, which is withdrawn after three days over inaccurate and racist outputs.
2017-01-01FAIR releases PyTorch, which becomes the dominant open-source deep learning framework industry-wide.
2013-12-09Facebook founds FAIR (Facebook AI Research) under Yann LeCun to pursue open, fundamental AI research.
Frequently Asked Questions
Who leads Meta AI today?
Alexandr Wang, former CEO of Scale AI, joined Meta in June 2025 as its first-ever Chief AI Officer and leads Meta Superintelligence Labs. Mark Zuckerberg remains Meta's overall CEO and has personally driven the AI reorganization.
What happened to Llama, Meta's open-source AI model?
Llama, released starting February 2023, made Meta the face of open-weights AI with over a billion downloads, but Llama 4's April 2025 launch drew criticism for lagging rivals. Meta has since pivoted to proprietary Muse models under Meta Superintelligence Labs, though it says it hopes to open-source future versions.
Why did Meta create Meta Superintelligence Labs?
After Llama 4's disappointing reception in mid-2025, Zuckerberg restructured Meta's AI efforts around Alexandr Wang, investing $14.3 billion in Scale AI to secure his leadership and consolidate FAIR, Llama teams and infrastructure into one unit aimed at achieving artificial general intelligence.
How much is Meta spending on AI infrastructure?
Meta has told investors it expects AI-related capital expenditures of roughly $125 billion to $145 billion in 2026, alongside a $100 billion multi-year AMD compute deal, reflecting an aggressive infrastructure buildout to support frontier model training.
Why did Yann LeCun leave Meta?
LeCun, Meta's longtime Chief AI Scientist and a Turing Award winner, departed in November 2025 after twelve years, reportedly over disagreements about the company's focus on large language models and its move away from open science, going on to found his own AI venture.
