Meta’s AI Overhaul Drives Up Security Incidents and Repair Costs

Meta is restructuring its workforce and deploying AI agents as it seeks to become an “AI-native” company, including recording employee workflows to generate training material for the agents. The strategy is intended to automate routine work and accelerate software development, but it has also raised questions about oversight of AI-generated code, cybersecurity accountability and whether headcount reductions are removing expertise needed to maintain critical systems.

Internal data cited in the latest report showed major technical and security incidents increased 40% after the adoption of AI-generated code, while the time employees spent repairing problems surged 70%. The figures suggest that productivity gains have yet to offset the additional burden of debugging, maintenance and incident response, challenging Meta’s effort to use artificial intelligence and workforce cuts to lower operating costs.