What 1500 Amazon engineers told company’s leadership over forced use of coding tool?

Amazon engineers are pushing back against a company policy favoring its AI coding assistant, Kiro, over superior third-party tools like Claude Code. Around 1,500 engineers have formally backed Claude Code, citing productivity concerns and questioning how sales teams can promote AWS Bedrock while restricted from using its offerings. This internal dissent highlights a growing demand for better AI tools.

Amazon‘s push to make employees use its in-house AI coding assistant Kiro over third-party tools like Anthropic’s Claude Code has sparked a quiet revolt inside the company. Around 1,500 engineers have formally endorsed the adoption of Claude Code through internal forums, pushing back against a policy they say is hurting productivity, Business Insider reported.The backlash centres on internal guidance rolled out late last year that steers teams toward Kiro for production code while restricting use of competing tools—including Claude Code, OpenAI’s Codex, and Cursor—without formal approval.

Amazon engineers say Kiro can’t keep up with rivals

The frustration runs deep. In internal discussion threads, as seen by Business Insider, employees argued that Claude Code simply outperforms Kiro, and that forcing adoption of a weaker tool slows down development.“A tool that can’t keep pace with rivals offers no real innovation,” one employee wrote, according to Business Insider. “And without competitive strength, Kiro’s only survival mechanism becomes forced adoption rather than genuine value.”The situation gets particularly awkward for engineers who sell AWS Bedrock—Amazon’s platform that offers customers access to third-party AI services, including Claude Code. Some questioned how they can credibly pitch a product they aren’t allowed to use themselves.“Customers will ask why they should trust or use a tool that we did not approve for internal use,” one sales engineer wrote internally.

Amazon says Kiro usage is growing fast

Amazon, for its part, says about 70% of its software engineers used Kiro at least once in January. A spokesperson told Business Insider the company is “seeing incredible improvements in efficiency and delivery from Kiro” and that customer growth is “rapidly accelerating.”The company maintains there’s no outright ban on Claude Code but applies “stricter requirements” for tools used in production code, with a process for seeking exceptions.An internal memo, first reported by Reuters in November, made the policy official. Signed by two senior VPs—Peter DeSantis and Dave Treadwell—it declared Kiro as Amazon’s “recommended AI-native development tool” and said the company does not plan to support additional third-party AI development tools going forward.

The $8 billion contradiction

What makes this especially complicated is Amazon’s own $8 billion investment in Anthropic, the company behind Claude Code. That stake is now reportedly worth over $60 billion. Anthropic, in turn, is a major AWS customer committed to using Amazon’s cloud infrastructure and Trainium chips.Some employees say the frustration deepened after Claude Code had apparently cleared security and legal review for production use—only for that language to be quietly edited and removed later.As one employee put it: “The trend is clear. More and more people would like to start using Claude Code and be officially supported by Amazon.”