Meta cuts 600 AI jobs from FAIR and other AI divisions

Meta is laying off about 600 employees from its legendary Fundamental AI Research (FAIR) group.

For years, FAIR has been behind foundational projects like PyTorch, the open source machine learning library used by millions, and groundbreaking work in Self-Supervised Learning (SSL) that taught computers to learn from massive amounts of unlabeled data. Other teams hit by the cuts include the product-related AI and AI infrastructure units, which get Meta's AI models into consumer hands.

In an internal memo seen by Axios, Meta's chief AI officer Alexandr Wang, stated the company hopes the reorganization will cut through bureaucracy. He said with a smaller team, "fewer conversations will be required to make a decision." Wang added that with the changes, "each person will be more load-bearing and have more scope and impact."

The one group not affected by the cuts is the new TBD Lab. This lab reportedly working on Llama 4.5 as a direct competitor to OpenAI's recently released GPT-5. To staff the lab, CEO Mark Zuckerberg launched an aggressive and very expensive hiring campaign, personally messaging top talent from rival companies like OpenAI, Google AI, and Google DeepMind. Wang, the Chief AI Officer, was appointed after Meta poured a whopping $14 billion into his former company, Scale AI.

Despite the star-studded recruiting drive, Meta has reportedly been struggling to keep its expensive new hires. High-profile incidents, like Shengjia Zhao, co-creator of ChatGPT, threatening to leave just days after joining. He only stayed after being given the title of "Chief AI Scientist".

Others simply walked away weeks after starting. Veteran employees are also reportedly not happy about the new dynamic, complaining about internal politics and unhealthy competition.

The employees affected by this latest round of layoffs are encouraged to apply for other positions inside the company, as their skills are needed in other divisions, according to Wang.