AWS AI coding tool decided to "delete and recreate" a customer-facing system, causing 13-hour outage, report says

Amazon's cloud division AWS experienced at least two outages involving its own AI tools, according to a Financial Times report. Amazon denies the connection and blames user error.

Amazon's cloud business AWS - which generates around 60 percent of the company's operating profits - suffered at least two outages in recent months where its own AI coding tools played a role, the Financial Times reports. Amazon rejects this characterization.

Four people familiar with the matter told the Financial Times that in mid-December, AWS experienced a 13-hour interruption to a customer-facing system after engineers allowed its Kiro AI coding tool to carry out certain changes. The agentic tool, which can take autonomous actions on behalf of users, decided the best course of action was to "delete and recreate the environment." Amazon posted an internal postmortem about the outage of AWS cost management features, according to the report.

Multiple Amazon employees also confirmed a second incident in recent months involving Amazon Q Developer. "We've already seen at least two production outages," one senior AWS employee told the newspaper. "The engineers let the AI agent resolve an issue without intervention. The outages were small but entirely foreseeable."

Amazon says misconfigured access controls caused the problem

Amazon firmly rejects this account. In an emailed statement to Reuters, an AWS spokesperson said: "This brief event was the result of user error - specifically misconfigured access controls - not AI."

Amazon described the December incident as an "extremely limited event." Only a single service, AWS Cost Explorer, was affected in one of two regions in mainland China. Compute, storage, database, and AI services were not impacted, the company said. The second incident did not affect a "customer facing AWS service," according to Amazon.

The company added that it was "a coincidence that AI tools were involved" and that "the same issue could occur with any developer tool or manual action." Amazon said it had not seen evidence that mistakes were more common with AI tools.

Regarding the December outage, Amazon told the Financial Times that the engineer involved had "broader permissions than expected - a user access control issue, not an AI autonomy issue." By default, Kiro "requests authorisation before taking any action," the company said.

AI tools had operator-level permissions with no peer review

The Financial Times reports, however, that the AI tools within AWS were treated as an extension of the operator and given the same permissions. In both cases, the engineers involved didn't need a second person's approval before making changes - something that would normally be required.

Following the December incident, AWS "implemented numerous safeguards," including mandatory peer review for production access and staff training. The fact that these measures were only introduced after the incidents sits uneasily with Amazon's claim that the problems were simply the result of user error.

Full statement by AWS

After publishing the original article, AWS contacted THE DECODER with a full statement:

“This brief event was the result of user error—specifically misconfigured access controls—not AI. The service interruption was an extremely limited event last year when a single service (AWS Cost Explorer—which helps customers visualize, understand, and manage AWS costs and usage over time) in one of our two Regions in Mainland China was affected. This event didn’t impact compute, storage, database, AI technologies, or any other of the hundreds of services that we run. Following these events, we implemented numerous additional safeguards, including mandatory peer review for production access. Kiro puts developers in control—users need to configure which actions Kiro can take, and by default, Kiro requests authorization before taking any action.”

AWS spokesperson

Update, Feb. 20, 2026: Added Amazon's full statement and additional context, including Amazon's position that it was "a coincidence that AI tools were involved" and that the internal postmortem characterized the event as an outage of AWS cost management features, not a broader AWS outage.