Amazon’s cloud division reportedly suffered at least one outage last year after an internal AI agent made a critical change to its system.
The incident caused a disruption lasting about 13 hours when the tool deleted and recreated part of its environment.
AWS said only one event affected customer-facing services and described it as user error caused by misconfigured access controls.
The company denied that artificial intelligence increases the risk of mistakes compared with human engineers.
The outages come as Amazon cuts thousands of jobs while expanding its use of AI.
Chief executive Andy Jassy has said automation will reduce routine work and shrink the workforce over time.
Security specialists questioned the company’s explanation.
They said AI systems can act quickly without fully understanding wider consequences.
Unlike humans, they may execute complex actions before an error becomes obvious.
AWS underpins large parts of the internet and many public services.
Previous outages have already highlighted the risks of relying on a small number of cloud providers.
Amazon said it has introduced extra safeguards, including mandatory peer review for production access.
It added that AI tools require human authorisation before taking major actions.
