Amazon Web Services, the cloud division of tech giant Amazon, recently responded to reports of AI-related outages with a detailed blog post correcting the information presented in a Financial Times report.

The Financial Times report had claimed that Amazon’s AI coding tools were responsible for at least two AWS outages in recent months, raising questions about the implications of using AI in cloud services.

In its blog post titled “Correcting the Financial Times report about AWS, Kiro, and AI,” Amazon clarified that the disruptions were due to user error in configuring access controls, not flaws in the AI tools themselves.

Amazon emphasized that the incidents did not impact customer-facing AWS services, challenging the definition of what constitutes an outage and highlighting the importance of accurate reporting.

The blog post refuted claims of a second event impacting AWS, asserting that the initial disruption was limited to a single service in one region and was quickly resolved without customer impact.

Despite the incident being classified as user error rather than AI error, Amazon acknowledged the need for continuous improvement in security and resilience measures to prevent similar disruptions in the future.

Amazon’s cloud division, AWS, is a key revenue generator for the company, with a significant focus on AI infrastructure and services. The company’s commitment to learning from operational experiences underscores its dedication to providing reliable cloud solutions to customers.

Following the incident, Amazon implemented new safeguards, including mandatory peer review for production access, to enhance security and prevent similar issues from occurring.

Overall, Amazon’s response to the AI outage reports reflects its proactive approach to addressing challenges in cloud services and ensuring the reliability of its offerings to customers.