Amazon’s cloud division reportedly suffered outages last year involving its own AI systems.
One disruption in December lasted 13 hours after an AI agent altered part of its environment.
Another incident in October affected websites for several hours.
AWS said the events were limited and blamed misconfigured access controls rather than artificial intelligence.
It added that only one case affected customer-facing services.
The company introduced extra safeguards, including mandatory peer review for production access.
The reports come as Amazon cuts thousands of jobs.
Chief executive Andy Jassy has said AI will improve efficiency but not replace staff directly.
He has also predicted that automation will reduce the workforce over time.
Some cybersecurity specialists question the company’s explanation.
They argue AI agents can act quickly without fully understanding wider consequences.
That lack of context can increase the risk of unexpected failures.
AWS remains a critical backbone for much of the internet and holds major public-sector contracts.
The incidents have renewed debate about reliance on a small number of cloud providers and the risks of rapid AI deployment.
