Hedgie@HedgieMarkets
🦔 Amazon is holding an engineering meeting Tuesday for a "deep dive" into a spate of outages, including incidents tied to AI coding tools. An internal memo seen by the Financial Times cited a "trend of incidents" with "high blast radius" and "Gen-AI assisted changes" among contributing factors, along with "novel GenAI usage for which best practices and safeguards are not yet fully established." Junior and mid-level engineers will now require senior sign-off on any AI-assisted changes.
This follows Amazon's website going down for nearly six hours earlier this month, and at least two AWS incidents linked to AI coding assistants, including the December outage where the Kiro AI tool decided to "delete and recreate the environment" during a fix attempt.
My Take
Amazon has been aggressively rolling out AI coding tools while cutting 16,000 corporate jobs in January, and now they're requiring senior engineers to review AI-assisted changes, which raises an obvious question about who's left to do that reviewing after the layoffs. Multiple engineers have reportedly said their teams are dealing with more "Sev2" incidents requiring rapid response since the cuts. Amazon disputed that headcount reductions caused the outages, but someone has to understand the systems well enough to catch AI mistakes before they take down the site for six hours, and that institutional knowledge walks out the door with every layoff.
I wrote about the AWS Kiro incident back in December when Amazon called it "user error" because the engineer gave the agent broader permissions than intended. But someone always has to decide what permissions to grant, and if AI behaves unexpectedly within those permissions, blaming humans doesn't prevent it from happening again. Now we're seeing the same dynamic play out across Amazon's retail infrastructure, and the solution is more human oversight of the AI tools that were supposed to reduce the need for humans.
Hedgie🤗