OpenAI and AWS expand partnership, easing enterprise AI deployment on Monday.com teams

OpenAI’s latest AWS move is less about model access than about removing the roadblocks that slow enterprise AI down. OpenAI models, Codex, and Amazon Bedrock Managed Agents were launched in limited preview on AWS on April 28, giving companies a way to build with frontier models inside the identity, security, procurement, and compliance systems they already use.

That matters for monday.com because so many enterprise buyers want AI to fit the stack they have already standardized on, not force a parallel platform. AWS said the new offerings carry enterprise controls such as IAM-based access management, AWS PrivateLink connectivity, guardrails, encryption at rest and in transit, CloudTrail logging, and integration with compliance frameworks. OpenAI also said eligible customers can apply Codex usage toward AWS cloud commitments, a detail procurement teams will notice fast because it turns an AI tool into something that can sit inside existing spend plans rather than outside them.

The Codex piece is especially relevant for engineering organizations. OpenAI said more than 4 million people use Codex every week, and described it as a tool for writing code, explaining systems, refactoring applications, generating tests, and modernizing legacy codebases. For teams at monday.com, where product velocity and reliability both matter, the signal is clear: coding agents are moving from side experiments into the same cloud environments where software is built, tested, and shipped. OpenAI said AWS customers with a cloud commitment and Bedrock access can start using Codex by configuring Amazon Bedrock as the provider, which lowers the setup burden for teams that already live in AWS.

The partnership also reflects a broader shift in how AI platforms are being bought. Enterprises increasingly want optionality across models, coding agents, and workflow agents without rebuilding their architecture around a single vendor. OpenAI and AWS said the goal is a single, consistent service with unified security, governance, and cost controls. AWS said the launch is in limited preview, but the direction is familiar to anyone trying to move AI from sandbox demos into production workflows: the winning product is the one security can approve, finance can charge back, and operations can monitor.

That lens fits monday.com’s own product direction. The company says it has more than 250,000 customers worldwide, and on March 11 it said it had built infrastructure that lets external AI agents sign up, authenticate, and operate directly inside the platform. For a work OS business, that is the difference between AI as a feature and AI as infrastructure. OpenAI and AWS are making the same bet for the cloud layer beneath it, and that is what will shape which teams can actually deploy AI at scale.