Amazon is pushing hard to make AI-assisted coding an everyday habit across the company. The target is clear—80 per cent of developers should be using AI for coding tasks at least once a week, according to the Financial Times. But the company isn’t leaving the choice of tool up to its engineers. It wants them on Kiro, its in-house AI coding assistant that launched in July 2025, and it’s actively discouraging everything else.An internal memo viewed by Reuters last November laid it out plainly: “We do not plan to support additional third-party AI development tools.” The memo, signed by two senior VPs—Peter DeSantis of AWS utility computing and Dave Treadwell of eCommerce Foundation—named Kiro as Amazon’s “recommended AI-native development tool.” OpenAI‘s Codex was flagged as “Do Not Use” after a six-month review. Anthropic’s Claude Code briefly got the same tag before the designation was reversed.A spokesperson told Business Insider that about 70 per cent of Amazon’s software engineers used Kiro at least once in January. The company wants to push that number higher—and fast.
Engineers aren’t thrilled—1,500 of them backed Claude Code instead
Not everyone is on board. Around 1,500 Amazon employees endorsed the formal adoption of Claude Code in an internal forum discussion, Business Insider reported. Many argued it was simply the better tool. “A tool that can’t keep pace with rivals offers no real innovation,” one employee wrote, adding that Kiro’s “only survival mechanism becomes forced adoption rather than genuine value.”The frustration is sharpest among AWS sales engineers who are tasked with selling Claude Code to customers through Amazon’s Bedrock platform—while being barred from using it for their own production work. “Customers will ask why they should trust or use a tool that we did not approve for internal use,” one of them wrote. Amazon maintains there’s no outright ban on Claude Code but says it applies “stricter requirements” for production tools.What makes this especially complicated is that Amazon has invested $8 billion in Anthropic, the company behind Claude Code, and signed a $38 billion cloud computing deal with OpenAI, whose Codex tool it also restricts internally.
Jassy’s bigger play: $200 billion in AI spending, 30,000 fewer jobs
The Kiro push is one piece of a much larger AI transformation under CEO Andy Jassy. He has told staff that AI-driven efficiency gains will shrink Amazon’s corporate workforce over time, and the numbers back that up. The company has cut 30,000 corporate roles since October 2025—the biggest reduction in its history—while committing a record $200 billion in capital expenditure this year, most of it directed at AI infrastructure and data centres, the Financial Times reported.Even as Amazon bets everything on its own AI tools, some of its own engineers aren’t convinced. Several told the Financial Times they preferred Anthropic’s Claude over Amazon’s Nova models for coding. One AWS engineer put it bluntly to FT: “I didn’t even know we had a model.”
