Sign up to see the future, today<\/p>\n
Can\u2019t-miss innovations from the bleeding edge of science and tech<\/p>\n
Even the gazillion-dollar empire that is Amazon is reeling from the costs of letting its employees run loose with AI agents.<\/p>\n
The Financial Times reports<\/a> that staff at the company have identified cases of \u201ccatastrophically expensive\u201d cost overruns as it switched over to using AI tools to perform what sound like pretty routine coding tasks.<\/p>\n In one case, Amazon blew $1.8 million by using Anthropic\u2019s Claude Sonnet AI to match author details with product listings on its namesake ecommerce website. This menial job somehow went 860 percent over the allocated budget. And equally alarming was that it took five months to notice the piles of money being set on fire.<\/p>\n \u201cIt\u2019s difficult to figure out how much anything [AI related] costs,\u201d one senior Amazon employee complained to the FT.<\/p>\n Senior engineers who presumably got their start before \u201cvibe-coding\u201d became all the rage warned that these weren\u2019t isolated incidents, lamenting that coding mistakes that were once \u201ctrivially cheap\u201d were now \u201ccatastrophically expensive.\u201d Amazon wasted $541,000 in unexpected costs related to building financial auditing tools, per the FT. And it also accidentally burned through $134,000 after giving AI the daunting task of improving delivery speeds across its logistics network.\u00a0(Magical thinking much? These things aren\u2019t genies, contrary to what Sam Altman promises<\/a>.)<\/p>\n Amazon\u2019s unintentional spending binges are part of a broader reckoning in the business world as leadership grapples with the costs of deploying AI tools across a company.\u00a0<\/p>\n Not long ago, the prevailing ethos of \u201ctokenmaxxing\u201d encouraged companies and software engineers to use AI agents, and especially coding tools, as much as possible. Inevitably, this led to unfortunate incidents such as one company that reportedly blew half a billion dollars in Claude usage fees<\/a> in a single month.<\/p>\n These costs concerns have been underscored by AI companies transitioning to token-based billing instead of flat subscription rates.<\/p>\n Amazon was certainly guilty of reveling in the debauchery \u201ctokenmaxxing\u201d era. It hosted an internal leaderboard, for example, that ranked employees based on how much they were using their AI tools. It was eventually shut down in May<\/a>, presumably as the spending hangover started to set in.<\/p>\n For AI to be burning a whole in its wallet is an ironic twist of fate for Amazon, because it\u2019s been culling its workforce for years to cut costs and pour more money into the tech. While conducting its latest round of layoffs affecting 16,000 employees<\/a>, it openly boasted of the \u201cefficiency gains\u201d from deploying AI across the company.<\/p>\n A spokesperson maintained that despite what it may look like, it\u2019s actually running a tight ship.<\/p>\n \u201cCherry-picking small, isolated examples where teams are learning from one another and portraying them as business as usual doesn\u2019t reflect how teams across Amazon are using AI,\u201d the spokesperson told the FT.<\/p>\n