Meta cuts 600 jobs from AI division during reorganization

Meta is cutting about 600 roles from its AI division as part of a reorganization, according to an internal memo seen by Axios.

The layoffs will affect the company’s FAIR research, product-related AI, and AI infrastructure teams, but will not impact the newer TBD Lab.

Meta employees in the US will be notified by 7am Pacific time on October 22 if their positions are affected.

The company will encourage those impacted to apply for other roles within Meta.

Meta continues to hire for its TBD Lab, which has recently added talent from OpenAI and Thinking Machines to its team.

The restructuring follows concerns from Meta leadership that previous AI efforts had become inefficient and were not delivering expected breakthroughs.

🔗 Source: Axios

🧠 Food for thought

Implications, context, and why it matters.

TBD Lab’s flat, ‘no deadlines’ approach sits alongside Meta’s push for speed

TBD Lab, Meta’s new AI research unit, runs with a very flat setup (few management layers) and no top-down deadlines since research timing is hard to predict 1. Zuckerberg called it a group science project 1.
The group has about 50 people 1 and sits alongside Meta’s push to operate more nimbly after the reorganization 2.
Layoffs in Fundamental AI Research (FAIR), product, and infrastructure teams hit groups that support model deployment with product integration, which could cause bottlenecks even if TBD Lab makes progress 2.
Meta scrapped its previous frontier model Behemoth and may make its new model closed instead of open source, which signals uncertainty beyond research 3.

FAIR and infrastructure veterans could bring hard-to-find AI implementation skills

About 600 laid-off employees came from FAIR research, AI product teams, and infrastructure divisions that support production AI models with consumer integrations 2.
Many veterans have shipped AI at Meta scale, which helps companies that prioritize implementation over frontier research.
Some employees enter a non-working notice period (paid but not actively working) with Nov. 21 termination dates and are encouraged to seek internal roles, creating a short window for outside recruiters 2.
An MIT report said about 95% of organizations see zero return on AI investments, so companies need people who turn research into revenue, which matches skills built in these product and infrastructure teams 3.