Amazon Convenes Emergency Meeting Following AI-Related Service Disruptions
Amazon convened an urgent internal meeting on Tuesday to address a concerning pattern of service outages linked to AI-assisted coding tools, raising questions about the reliability of artificial intelligence in critical infrastructure management.
Reports indicate that Amazon’s retail technology executive, Dave Treadwell, held a mandatory “deep dive” meeting to discuss multiple recent incidents where generative AI-assisted code changes contributed to system failures. The company experienced a six-hour e-commerce outage last week that prevented customers from checking out and accessing account information, which Amazon attributed to “software code deployment.”
Amazon acknowledged that “best practices and safeguards” around generative AI usage haven’t been fully established yet. The company is implementing new safety measures requiring additional senior review of AI-assisted production changes. Junior and mid-level engineers will now need sign-off from senior managers before pushing AI-generated code to production systems.
Separately, Amazon Web Services reportedly experienced at least two outages in recent months involving its Kiro AI coding tool, including a 13-hour disruption in December. While Amazon insists these were “user error, not AI error,” the incidents highlight growing concerns about rapidly deploying AI automation in mission-critical systems as tech companies race to integrate the technology.
Source: CNBC