Amazon convened an emergency meeting after a series of site and infrastructure outages traced to code changes made with AI tools, whose best practices and safety measures the company admits are still evolving. In December, developers followed an AI agent Kiro’s suggestion to delete and rebuild the AWS environment, and this month a six‑hour outage left key functions offline after a buggy deployment. Senior VP David Trevorrow now requires approval from senior and middle‑level developers for any AI‑generated changes. Engineers say the spike in incidents coincides with the January layoff of 16,000 employees, a claim Amazon denies.

Financial Times