tech
AI agents meant to replace Meta workers made “large-scale, disruptive actions”
Report shows Meta’s challenges replacing people with AI agents.
TL;DR
- Meta explored a plan, codenamed Project OT, to make the company 'AI native' by reducing some teams by up to 60% and using AI agents for daily tasks.
- The plan involved scenario planning for layoffs and redeployments, with potential headcount reductions of 25% or more.
- Savings from potential layoffs were intended to fund high-performing employees, especially in AI engineering.
- An internal document described 'AI native' as a company where AI-ready tools and agents interact, workflows are automated, and new builds are AI-first.
- Meta's CTO reported a significant increase in code changes to internal software platforms but a smaller increase in new or upgraded features for users.
- Internal posts indicated AI agents made disruptive actions, leading to a 40% increase in major technical and security incidents and a 70% increase in employee time spent resolving them.
- Meta CEO Mark Zuckerberg reportedly canceled the second wave of layoffs planned for November and acknowledged that AI agent development had not accelerated as expected.
- The company paused a program where it tracked employees' keyboard and mouse input to train AI agents.