OpenAI has agreed to acquire Promptfoo, a startup that builds security tools for large language models, as it pushes deeper into enterprise AI agents and the systems that run them. The deal shows how seriously AI companies now take security as more businesses use agents to handle sensitive workflows, internal data, and automated tasks that need tighter control.
Promptfoo, founded in 2024 by Ian Webster and Michael D’Angelo, built tools that help companies find weaknesses in AI systems before those issues cause real damage. Its products focus on red-teaming, testing for prompt injections, spotting data leaks, and checking whether AI tools stay within policy. OpenAI plans to add those capabilities to OpenAI Frontier, its platform for building and running AI coworkers.
In a blog post Monday, OpenAI said the acquisition will bring Promptfoo’s technology directly into Frontier so enterprises can test agent behavior, monitor risk, and support compliance needs more easily. The company also said it plans to keep developing Promptfoo’s open-source project, which already has wide use among large businesses.
Why this deal matters
The move comes at a time when companies want AI agents to do more real work, but they also need stronger guardrails before deployment. Promptfoo says more than 25% of Fortune 500 companies use its tools, which helps explain why OpenAI sees it as a strong fit for enterprise customers.
“Automated security testing and red-teaming capabilities will become a native part of the Frontier platform,” OpenAI said.
That line captures the heart of the deal. OpenAI wants “native” security inside the platform, not as an extra layer added later.
“Joining OpenAI lets us accelerate this work,” said Ian Webster, co-founder and CEO of Promptfoo.
OpenAI did not disclose the value of the transaction, and the deal still needs to meet standard closing conditions.
