San Francisco: Amazon is aggressively steering its internal engineering teams toward its proprietary AI coding assistant, Kiro, according to a memo reviewed by Reuters. The internal directive, signed by Peter DeSantis (SVP, AWS Utility Computing) and Dave Treadwell (SVP, eCommerce Foundation), urges employees to preferentially use Kiro over external AI-powered development tools.
In the memo, Amazon acknowledges the continued use of existing third-party tools but makes it clear it will not support any additional external AI development platforms going forward. “While we continue to support existing tools … we do not plan to support additional third-party, AI development tools,” it states.
The company is positioning Kiro as its recommended “AI-native” development tool, saying that feedback from its internal “builder community” is critical to refining the tool. The memo explicitly discourages the use of other tools from firms like OpenAI (Codex), Anthropic (Claude Code), and start-ups such as Cursor.
Amazon first launched Kiro in July 2025. It is an AI-powered code-generation tool that can build apps and websites purely from natural-language instructions. While Kiro draws on technology from Anthropic, it is not simply a rebrand of Claude Code, the memo notes.
The timing of this internal push is politically and strategically significant. Amazon has made substantial investments in AI companies: it has reportedly poured around US$ 8 billion into Anthropic, and locked in a seven-year, US$ 38 billion cloud-services contract with OpenAI. Despite these ties, the memo makes clear that Amazon wants to reduce its internal reliance on these very players when it comes to its own coding workflows.
The move also comes as Amazon expands the reach of Kiro, having recently made it available globally with new features, according to the same memo.
Industry observers see this as a dual-pronged strategy: first, to consolidate control over its AI development ecosystem, reducing reliance on third-party tools; and second, to accelerate internal innovation by tightening the feedback loop between Amazon’s engineers and its own AI development tool. By asking engineers to “help shape” Kiro aggressively, Amazon may be hoping to rapidly iterate and improve its performance, making it more competitive with external offerings.
However, the mandate carries risks. Critics note that restricting access to external tools could stifle experimentation and cross-pollination of ideas. Many developers thrive by mixing and matching tools; forcing a lock-in to Kiro could potentially undermine innovation if Kiro does not keep pace with the rapid advancements elsewhere.
Furthermore, the internal memo reflects Amazon’s long-term ambition it may not just want Kiro for internal use, but could be preparing to position it as a broader commercial product, perhaps as part of its AWS offering. If successful, Kiro could become a serious competitor in the AI code-generation space, challenging tools built by OpenAI, Anthropic, and other startups.
In short, Amazon’s internal call to favor Kiro signals a major shift in its AI strategy: building from within, consolidating its tech stack, and betting heavily on its homegrown platform to power its future software development even as it continues to invest deeply in external AI players.