Amazon AI guidelines are urging software engineers to be selective about using artificial intelligence, warning that measuring success by token consumption or code output can increase costs and complexity without demonstrating useful results.

Key Takeaways

  • Amazon published the memo on October 27, 2026, after circulating it internally in September.
  • The company has not imposed caps on employee token usage, a spokesperson said.
  • Amazon says AI activity metrics should be considered alongside other data, with an emphasis on long-term outcomes.

Amazon AI guidelines favor outcomes over activity

The memo revises how Amazon defines the roles of software development engineers and managers as AI becomes part of their work. It discourages tokenmaxxing, a practice in which developers use AI as much as possible with little regard for the output.

Amazon described token usage, adoption rates and measures of code changes or reviews as deeply flawed ways to assess AI use across teams or organizations. It did not say to disregard those measures entirely, but urged employees to consider token usage and code generation with care and judgment alongside other data.

The company also said engineers should not feel obliged to use AI for every task. Identifying work that does not benefit from current AI tools demonstrates the judgment Amazon is asking for, the memo said.

Productivity and the cost of AI tools

Tokens are units of data processed and generated by AI models. Some companies have limited employee token use or switched to cheaper open-source models for certain tasks as AI bills rise. Amazon has not instituted token caps, its spokesperson said.

In May 2026, Amazon shut down an internal leaderboard after finding that employees were tokenmaxxing to climb its rankings.

Marc Brooker, a vice president and distinguished engineer at Amazon Web Services, said the company has considered measures such as feature velocity, including the number of launches and successful production changes, when assessing engineering productivity. He said Amazon also needs room to experiment while avoiding spending tokens on problems that do not matter.

Read More News

More FinanceVelix news

Disclaimer

This FinanceVelix article provides general information, not personalized financial, investment, legal, or tax advice. Market prices, interest rates and economic conditions can change; investments carry a risk of loss. Facts reflect the material available when published and may change. Check important information independently and consult a qualified professional when needed. Mentioning an asset, company, or service is not an endorsement.