In December 2025, Amazon Web Services (AWS) faced a significant 13-hour disruption attributed to human error rather than AI malfunction, as clarified by an AWS spokesperson. The incident, impacting a service in Mainland China, reportedly occurred after engineers allowed the Kiro AI system to execute tasks that led to a critical environment reset. While the Financial Times highlighted potential AI involvement, AWS emphasized that the issue stemmed from misconfigured access controls. This event was described as limited, with no effect on core services or customer experience, contrasting sharply with a more extensive outage experienced in October. AWS has since reinforced its protocols by implementing additional safeguards, including mandatory peer reviews and enhanced AI training. As industries increasingly rely on cloud providers like AWS, experts note that even minor outages can lead to widespread implications across the internet, raising questions about the frequency and impact of these disruptions.
Source link
Share
Read more