OpenAI acquires Promptfoo

OpenAI announced in March 2026 that it had agreed to acquire Promptfoo, a security startup founded in 2024 to help companies test large language models and AI agents against adversarial prompts and other attacks. Promptfoo’s open-source and commercial tools, which TechCrunch reported were used by more than a quarter of Fortune 500 companies, had previously raised about $23 million and were valued at $86 million after a funding round in mid-2025.

OpenAI did not disclose the terms of the deal. The company said Promptfoo’s technology would be integrated into OpenAI Frontier, its enterprise platform for AI agents, to provide automated red-teaming, security evaluation of agentic workflows, and ongoing monitoring for risk and compliance issues, while continuing to develop Promptfoo’s open-source offering separately.

The acquisition was one of several security-focused moves by OpenAI in the period, alongside a research preview of a separate code-vulnerability tool, Codex Security, and reflected a broader shift among frontier labs toward acquiring or building dedicated evaluation and red-teaming infrastructure as enterprise customers increasingly asked for evidence that agentic AI deployments had been adversarially tested before going into production.