OpenAI Acquires Security Firm Promptfoo to Fortify AI Agent Safety

OpenAI announced on Monday its acquisition of cybersecurity startup Promptfoo. The move aims to integrate Promptfoo's specialized testing and security tools directly into OpenAI's Frontier platform, which is designed for developing AI agents.

The financial terms were not disclosed. Promptfoo's entire 11-person team will join OpenAI. The startup's technology focuses on validating and securing complex AI systems, a growing concern as these agents increasingly interact with live data and critical infrastructure.

"Validating and securing AI agents is now one of the most significant challenges in the field," said Promptfoo CEO Ian Webster. "By joining OpenAI, we can scale our work faster and make robust safety tools more accessible to developers building practical AI applications."

OpenAI confirmed it will maintain and develop Promptfoo's open-source project, a toolkit that allows engineers to test prompts and compare outputs across different AI models, including GPT, Claude, and Gemini.

Promptfoo had recently established its footing, closing an $18.4 million Series A round led by Insight Partners in July 2025, with participation from Andreessen Horowitz. According to PitchBook data, the startup had raised a total of $22.68 million and was valued at $85.5 million at that time.

The acquisition highlights the intensifying focus on AI security. It also follows a period of significant fundraising by investor Andreessen Horowitz, which announced a $15 billion fund in January with portions dedicated to growth, apps, and infrastructure ventures.

Source: CNBC