 |
VOOZH | about |
|
VOOZH | about |
We’re so glad you’re here. You can expect all the best TNS content to arrive Monday through Friday to keep you on top of the news and at the top of your game.
Check your inbox for a confirmation email where you can adjust your preferences and even join additional groups.
Follow TNS on your favorite social media networks.
Become a TNS follower on LinkedIn.
Check out the latest featured and trending stories while you wait for your first TNS newsletter.
Amazon has given its estimated tens of thousands of developers immediate access to Anthropic’s Claude Code, and they’ll soon have access to OpenAI’s Codex, opening up agentic coding tools beyond its own Kiro.
In an internal staff note seen by Business Insider, VP of Amazon Software Builder Experience Jim Haughwout said both tools will run on Amazon Web Services and Amazon Bedrock, Amazon’s fully managed service for building and scaling generative AI applications.
Doing so will allow the e-commerce juggernaut to give its developers access to top-tier coding assistants while avoiding complex infrastructure setup, simplifying capacity management, and tightly controlling data security and compliance.
In his note to Amazon staff, Haughwout says the technology company is opening up access to Claude Code and Codex on May 12 to ultimately better serve its customers: “To help you invent more for customers, we are expanding the agentic AI tool available to you.”
But it seems Amazon may finally be caving to its developers’ demands after hearing their complaints about restrictions on third-party tools and the company’s push to use its own Kiro.
Launched in 2025, Kiro is an agentic coding service built on Bedrock. As Business Insider reported in February, internal messages from Amazon told its team to prioritize Kiro for production code over other non-approved third-party tools, which included Claude Code. But Amazon employees were swift to bite back, with one discussion thread reportedly containing about 1,500 employee endorsements urging the company to formally adopt Claude Code.
Just a few months later, that approval has come.
Claude Code, and soon Codex, is now approved for production use, and employees no longer need special clearance to use the tool.
Still, as an Amazon spokesperson told Business Insider, Kiro remains: “Our builders are using Kiro for agentic coding, and now with both Claude Code and Codex running on AWS, we are making additional tools available as well.”
In making Claude Code and Codex available across the company, Amazon is significantly expanding its use of third-party coding tools.
“Constraint now needs to shift to review, validation, and making sure the system behaves the way you think it does.”
When asked what this likely means for the technology company’s developers, Satyam Dhar, staff software engineer at Galileo, formerly at Adobe and Amazon, tells The New Stack that he expects the biggest impacts will be on internal workflows: “Constraint now needs to shift to review, validation, and making sure the system behaves the way you think it does.” Meanwhile, leadership, he predicts, will have “to rethink how they evaluate productivity and output.”
Amazon’s new approval for Claude Code and Codex marks a continued shift towards AI- and agentic-led work in e-commerce.
For example, in a blog post last spring, Walmart detailed its own development of purpose-built agentic AI tools.
The retailer highlighted its “surgical” approach to agentic AI, saying it prioritizes work on agents for retail-specific tasks rather than for multiple use cases. Its GenAI-powered shopping assistant, Customer Assistant, and Trend-to-Product are just a few tools among the examples the retailer cites as “early success and meaningful impact of agentic AI.”
Amazon has already made similar forays into AI-powered e-commerce, such as the 2024 launch of Rufus, its AI shopping assistant. But by approving company-wide use of Claude Code and Codex, it’s giving developers wider access to tools that could potentially help it develop and ship more new features faster.
Walmart, for its part, is no stranger to AI coding assistance. In a Q4 fiscal 2025 earnings call, the retailer said AI-powered coding assistance helped save 4 million developer hours, as reported by CIO Dive. As told to the Wall Street Journal at the end of last year by Sravana Karnati, Walmart’s executive vice president of global technology platforms, over 95% of Walmart’s engineers have already been using AI coding assistant tools.
When asked how Amazon’s move will shape software development, Dhar thinks it’s a signal that AI coding is moving beyond experimentation to start becoming the production norm:
“It is changing where engineers spend their time. Less writing code line by line, but more time defining structure, reviewing outputs, and thinking through system behavior.”
Amazon’s update comes as the company continues to deepen its ties with both Anthropic and OpenAI.
In April, Amazon and Anthropic announced a deal under which Anthropic makes the Claude Platform available on AWS, accepts up to $25 billion in investment from Amazon, and commits to investing $100 billion in AWS technologies over 10 years to secure up to 5 gigawatts (GW) of Amazon’s Trainium and Graviton cores.
Earlier in the year, Amazon inked a deal with OpenAI, investing up to $50 billion in the AI company as the two commit to co-create a Stateful Runtime Environment powered by OpenAI models, available on Amazon Bedrock for AWS customers.
