Eligible OpenAI enterprise customers can use part of their existing commitments to buy CrowdStrike's Falcon platform, as the two companies expand their work on securing AI agents. 

[For more news, click here]

Companies are putting AI into more parts of their businesses, and some of those systems are being given access to code, data and internal tools. That makes security part of the decision from the start, particularly when an AI agent can carry out tasks instead of waiting for a person to handle each step. The way enterprises pay for those security controls is starting to change too, with AI providers creating closer links between their own commercial agreements and the software customers use alongside them. 

CrowdStrike is bringing its Falcon platform to the OpenAI Marketplace, giving eligible OpenAI enterprise customers the option to use part of their existing OpenAI commitments toward qualifying CrowdStrike purchases. CrowdStrike announced the move, adding a new purchasing route to a partnership that has also grown around the security of AI agents and the use of OpenAI's cyber reasoning capabilities in CrowdStrike's security platform. 

OpenAI's Marketplace does not allow customers to spend their entire commitment freely on third-party software. Eligibility is decided for each customer, product and purchase, and the customer contracts with and pays the participating partner directly. The qualifying amount is then applied against the customer's OpenAI commitment. For companies already spending heavily on OpenAI, that gives them a way to put some of the money they have committed toward security software as their AI deployments grow. 

CrowdStrike is building security around Codex agents 

The Marketplace announcement follows an expansion of the two companies' technical partnership. Falcon Guardian extends runtime protection to supported OpenAI Codex agents, giving security teams visibility into agent activity, detecting compromised or unauthorized behaviour and enforcing runtime controls. CrowdStrike is also bringing OpenAI's GPT-5.6 Cyber into the Falcon platform, starting with its Frontier AI Readiness and Resilience, or FAIRR, Service before expanding the capability across Falcon. 

Falcon Guardian is focused on what an agent does while it is running, while GPT-5.6 Cyber is being used to add advanced cyber reasoning to CrowdStrike's security work. As agents are given access to business systems, security teams need to keep track of what they can reach and what they are doing. OpenAI has also been adding permissions, identity, monitoring and governance controls to its enterprise agent infrastructure as it moves toward wider use of agents in business settings. 

“CrowdStrike and OpenAI are bringing together the best of cybersecurity and frontier AI to secure the agentic enterprise,” said Daniel Bernard, chief business officer at CrowdStrike. “The OpenAI Marketplace gives customers a powerful new way to apply their AI investments toward cyber defense, bringing the Falcon platform and OpenAI’s frontier intelligence together to drive faster outcomes for defenders.” 

For enterprise buyers, the purchasing model is worth watching because AI and security have usually been handled as separate technology decisions. A company deploying an AI agent may already have an OpenAI commitment in place and then need additional software to monitor or control what that agent can do. Allowing part of that commitment to be used for an eligible security product gives procurement teams another way to handle the two purchases, even though they remain separate agreements. 

“Businesses trust OpenAI for frontier intelligence, and they want that same intelligence in tools they use every day. With the OpenAI Marketplace, enterprise customers can now use their existing OpenAI commitments to access solutions built by partners like CrowdStrike on OpenAI’s models and platform,” said Brian Landsman, VP, Global Partnerships at OpenAI. “Together, we’re combining our frontier intelligence with CrowdStrike’s platform to give customers more choice in how they put AI to work.” 

AI agents are also creating new questions around access inside corporate systems. Unlike traditional software tied to a specific employee or application, an agent can operate with permissions that allow it to access data, use tools and carry out tasks on its own. Security teams therefore need to know what an agent is allowed to access, what actions it is taking and whether that behaviour changes from what was expected. Those concerns are becoming more relevant as companies move agents from testing into real business workflows. 

The Marketplace gives eligible customers a way to buy Falcon using part of an existing OpenAI commitment, bringing the purchasing decision closer to the AI deployment itself. There are still limits to what the arrangement changes for customers. An OpenAI commitment cannot simply be converted into a general cybersecurity budget, and buying Falcon does not determine how an enterprise should set permissions or govern its AI agents. The value for customers is in having another way to bring security spending into an AI project that may already have an approved budget. 

As more companies give AI systems access to the tools and information they use to run their businesses, security will increasingly have to be considered alongside the initial AI investment. CrowdStrike's Marketplace availability gives eligible OpenAI customers another way to make that connection, while the wider partnership puts the companies' technology together across agent protection, cyber reasoning and enterprise procurement. 

Related Articles 

VAST Data and CrowdStrike Extend Cybersecurity to Enterprise AI Data 

With AI Agents Up 460% in a Year, BeyondTrust Launches a Governance Tool to Match 

Tenable and Anthropic Think Agentic AI Can Finally Close the Gap Between Cyber Discovery and Response