Amazon’s cloud computing arm suffered two outages caused by its own artificial intelligence tools. The Financial Times reported that a 13-hour outage in December was triggered by an AI agent, Kiro, which “deleted and then recreated” part of its environment.
The incident is the latest to raise questions about Amazon’s reliance on AI as it lays off human employees. In January, the company confirmed plans to cut 16,000 jobs. CEO Andy Jassy had previously stated that efficiency gains from AI would lead to workforce reductions.
Experts say that while engineering errors caused by humans and traditional tools are common, AI-related mistakes can be more significant due to their lack of contextual understanding. Security researcher Jamieson O’Reilly said that AI agents often run in isolated environments and cannot fully grasp the consequences of their actions.
Amazon has denied that its AI tools were responsible for the outages, attributing them to “user error” and misconfigured access controls. However, experts believe that it’s impossible for Amazon to completely prevent internal AI agents from making errors in the future due to the complexity of these systems.
Source: https://www.theguardian.com/technology/2026/feb/20/amazon-cloud-outages-ai-tools-amazon-web-services-aws