Key Takeaways:
Meta has reassigned approximately 7,000 employees to new artificial intelligence units, a significant internal restructuring aimed at closing the performance gap with rivals like OpenAI and Google. The move, internally dubbed the "draft," is part of CEO Mark Zuckerberg's broader strategy to accelerate the company's AI development.
"I think that this is going to be a very big advantage, if we can do it," Zuckerberg said in a leaked audio recording from an internal meeting, referring to using Meta's own employees for training AI. He added that the average Meta employee has a "significantly higher" intelligence than the contractors typically used for data labeling.
The reassignments have funneled employees into several new groups, including a central unit called Applied AI (AAI) and more specialized teams like "Agent Transformation Accelerator" and "Agent Data and Optimization." For the thousands of employees selected based on "strong performance," the move was non-negotiable, leading to a mix of relief and uncertainty amid wider company layoffs.
This aggressive pivot to in-house AI training represents a core hypothesis for Meta: that its existing, high-skilled workforce can provide a crucial data advantage to train its models more effectively than the armies of contractors used by competitors. While Meta's AI models have trailed rivals in performance, this internal mobilization could allow the company to more rapidly improve its technology by using its own top-tier talent for the intensive work of data labeling and model optimization.
The internal reaction to the "draft" has been mixed, with some employees expressing relief at being spared from layoffs, while others are uncertain about their new roles. On anonymous workplace app Blind, some employees reassigned to the AAI group questioned its function, with two Meta employees telling Business Insider they expect the work to involve data labeling—the process of manually tagging data to train AI models.
The strategy appears to leverage the experience of AI chief Alexandr Wang, co-founder of Scale AI, a major data-labeling company. This move aligns with Zuckerberg's stated belief that Meta's own employees can provide higher-quality data for training its AI.
Meta's all-in push on AI, which includes reassigning thousands of its employees, is a high-stakes bet to catch up with competitors. For investors, this signals a potential long-term advantage if the company can successfully leverage its internal talent to create superior AI models. However, it also raises questions about the immediate impact on productivity and morale. Meta's stock, currently trading at a forward P/E ratio of around 20x, could see a positive re-rating if this strategy translates into demonstrable AI progress and new revenue streams. The success of this internal "draft" will be a key factor for investors to watch in the coming quarters.
This article is for informational purposes only and does not constitute investment advice.
