Amazon’s cloud division suffered at least one outage last year involving its own AI systems, according to reports.
An incident in December lasted about 13 hours after an AI agent altered part of its operating environment.
AWS said the disruption was limited and blamed misconfigured access controls rather than the AI itself.
The company insisted there was no evidence that AI tools cause more errors than human engineers.
The episode has raised questions about Amazon’s rapid expansion of AI while cutting jobs.
The group announced 16,000 layoffs in January after 14,000 earlier reductions.
Chief executive Andy Jassy has said AI will improve efficiency but not simply replace staff.
Cybersecurity experts argue AI agents can act without understanding wider consequences.
They say these systems require constant human oversight and clear limits.
Complex behaviour makes future mistakes difficult to eliminate completely.
AWS runs critical infrastructure for large parts of the internet.
Previous outages have already highlighted the risks of relying on a small number of major cloud providers.
