Amazon AI Projects Over Budget by Up to $1.8 Million

Amazon has reported multiple instances of severe budget overruns in its AI projects. One project, which aimed to use Claude Sonnet to match author information on product pages, ultimately failed, costing $1.8 million—860% over budget—and was only discovered five months later. Another financial auditing tool project unexpectedly overspent by approximately $541,000, while a project intended to optimize logistics delivery speed misallocated $134,000, with both issues identified over two weeks later. Amazon engineers pointed out that the problems stemmed from deployment errors and a lack of cost controls, noting that small mistakes in traditional systems can lead to hefty bills in AI projects. The company is developing automated measures to prevent similar situations from occurring in the future. Amazon stated that these cases involve only a few teams and do not represent the overall use of AI. Previously, Amazon had shut down its internal AI usage leaderboard due to employees having Agents perform meaningless tasks to climb the rankings, resulting in excessive Token consumption.