Meta has abandoned Project OT, an initiative launched by Mark Zuckerberg in January that aimed to replace human workers with AI across the company. The plan had targeted cuts of up to 60% in some teams, but was shelved after the AI models failed to deliver expected productivity gains.
Technical and security incidents became significantly more frequent under the AI deployment. According to Reuters, these incidents occurred 40% more often compared to the previous year, and the time required to contain them increased by 70%. The combination of lower-than-expected performance and mounting operational problems made the cost-benefit case for AI replacement untenable.