Amazon’s cloud division suffered outages last year that were reportedly triggered by its own AI systems.
One disruption in December lasted about 13 hours after an AI agent automatically deleted and rebuilt part of its environment.
The incident affected only limited customer services, according to the company.
Amazon said the problem stemmed from user misconfiguration, not the AI itself.
It added that new safeguards and mandatory peer reviews have since been introduced.
AWS underpins large parts of the internet, so outages raise wider concerns about reliance on a few providers.
A separate failure in October temporarily knocked dozens of websites offline.
The reports come as Andy Jassy oversees major job cuts.
Amazon announced 16,000 layoffs in January after 14,000 reductions last year.
The company says the cuts relate to efficiency and culture rather than replacing staff with AI.
Some cybersecurity experts dispute Amazon’s explanation.
They argue AI agents can act without fully understanding wider system risks.
Unlike human engineers, they may execute harmful actions quickly without recognising the consequences.
Amazon maintains the outage was a brief and limited event and not caused by AI failure.
