Amazon Engineers Move to Cap AI Spending After Cost Overruns
Amazon’s (AMZN) engineering teams have found cases where moving work from hand-written code to AI models blew through project budgets, including $1.8 million spent running Anthropic’s Claude Sonnet on a job that never shipped. Senior engineers laid the cases out for staff at an internal meeting on July 28, 2026 and described the results as “catastrophically expensive,” the Financial Times reported, citing multiple people familiar with the presentation. They told colleagues they are now building automated guardrails to cap what future projects can spend.
The largest example matched author records against product listings on Amazon’s retail site. Spending came in 860% above what that project had budgeted, took five months to surface, and the deployment failed anyway. Two smaller cases were shown alongside it: about 134,000 on work to improve delivery speeds across Amazon’s logistics network, caught after more than two weeks.
Amazon said it is “experimenting, learning and improving” how it uses the technology, including how it drives cost efficiencies. The company also said that presenting a handful of isolated examples as business as usual misrepresents how its teams work with AI, and that the cases covered a small number of groups inside a corporate workforce of roughly 300,000.
Token billing dynamics
Staff were told that mistakes costing almost nothing in conventional systems turn expensive once a model does the work. Anthropic charges 15 per million output tokens for Claude Sonnet 4.5 and 4.6, with Sonnet 5 at an introductory 10 until August 31, 2026. At Sonnet’s standard input price, $1.8 million buys on the order of 600 billion input tokens.
A retry loop that re-sends the same context, or a batch job pointed at an entire catalogue instead of a sample, throws no exception — it produces an invoice on a monthly billing cycle. Anthropic’s documentation notes Claude 4.7 and later use a newer tokenizer producing roughly 30% more tokens for the same text.
Controls and prior context
AWS Bedrock offers batch inference (half on-demand), Flex/Priority tiers, and Intelligent Prompt Routing. Anthropic cache reads are a tenth of standard input rate; Haiku 4.5 is roughly a third of Sonnet rates.
Earlier in 2026 Amazon shut an internal leaderboard ranking employees’ use of its Kiro developer platform after staff inflated token consumption (“tokenmaxxing”). Automated guardrails are the next move, putting budget enforcement into the deployment path.
Amazon is expected to spend about $200 billion in capital expenditure across 2026, most of it on AI and data-center infrastructure. The company reports second-quarter results after the close on July 30, 2026.