Code Error at Amazon Causes Disruption in E-Commerce Infrastructure
During a meeting held by Amazon engineers, code written for testing purposes accidentally affected the live system and caused disruptions in some e-commerce services.
It was stated that some of the outages were linked to Amazon’s artificial intelligence coding assistant Q. Following the outage, a 90-day code security reset was initiated across critical engineering systems. The disruption was detected and resolved in a short time. However, the incident once again demonstrated that even a small code change on large-scale e-commerce platforms can affect millions of users.
Amazon is preparing to launch a comprehensive internal review regarding artificial intelligence–assisted software development processes and potential system outages that have recently come to the agenda. The company’s engineering teams and senior executives are planning a series of meetings, particularly to address the impact of AI-assisted code generation on system security and operational continuity.
Code Generation Tools Are Increasing at Amazon
In recent years, Amazon has rapidly increased the use of artificial intelligence–based code generation tools in its software development processes. While these tools help developers write code more quickly, some experts point out that automatically generated code could pose risks if it is integrated into live systems without sufficient testing.
At the center of the discussions are certain technical outages experienced within Amazon’s infrastructure and the question of whether these outages are related to AI-assisted development processes. During internal meetings, engineers are expected to examine the possible causes of system disruptions in detail and conduct evaluations particularly on testing processes, code verification mechanisms, and the security of automation tools.
Millions of Sellers and Hundreds of Millions of Customers in the E-Commerce Infrastructure
Amazon’s technology infrastructure supports a massive e-commerce ecosystem that serves millions of sellers and hundreds of millions of customers. For this reason, even a small software error can create impacts on a global scale. Therefore, the company aims to establish stronger control mechanisms and security layers in artificial intelligence–assisted software development processes.
According to experts, although code generation with artificial intelligence accelerates software development processes, human oversight, comprehensive testing procedures, and secure deployment policies remain critically important. It is stated that the evaluation process initiated by Amazon could serve as an important reference for how AI-assisted software development practices should be managed across the technology sector.
Amazon: It Was Not Caused by Code Written by AI
Amazon stated that some claims reported in the media do not fully reflect the reality and said that the comments suggesting the outages were caused by code written by artificial intelligence are not correct. The company emphasized that it continuously improves infrastructure security and system resilience and that its engineering teams regularly review system performance.