OpenAI has announced its intention to acquire Promptfoo, an AI safety platform that assists enterprises in identifying and fixing vulnerabilities within AI systems during the development phase. This acquisition is aimed at accelerating the intelligent agent security testing and assessment capabilities of OpenAI Frontier, the platform used for building and operating AI collaborative partners.
As AI collaborative partners are deployed into real business processes, assessment, security, and compliance have become fundamental requirements. Enterprises require systematic methods to test agent behavior, uncover risks before deployment, and maintain clear records to support long-term supervision, governance, and accountability.
The Promptfoo team, led by Ian Webster and Michael D’Angelo, has built a robust set of tools that have earned the trust of over 25% of Fortune 500 companies. They also maintain a widely used open-source command-line interface (CLI) and libraries for evaluating and red team testing of large language model (LLM) applications. OpenAI plans to continue building on this open-source project while enhancing the enterprise-level capabilities integrated within Frontier.
Key Capabilities to be Integrated
Platform-Built Security and Assurance Testing: Automation of security testing and red team testing capabilities will become a native part of the Frontier platform, helping enterprises identify and fix various risks, including prompt injections, jailbreaks, data leaks, tool misuse, and agent behaviors outside of policy boundaries.
Security and Assessment Integrated into Development Workflow: Frontier will deeply integrate related workflows to identify, investigate, and fix agent risks earlier, making security a core part of enterprise AI system development and operation.
Supervision and Accountability: Integrated reporting and traceability capabilities will help organizations document testing processes, monitor long-term changes, and meet the growing requirements for AI governance, risk management, and compliance (GRC).
“Promptfoo was born out of an urgent need to equip developers with practical capabilities to safeguard AI security. As AI agents become more closely connected to real data and systems, securing and verifying them becomes more challenging and critical than ever. Joining OpenAI allows us to accelerate our efforts to provide stronger security and governance capabilities for teams building real-world AI systems.” — Ian Webster, Co-founder and CEO of Promptfoo
OpenAI warmly welcomes the Promptfoo team and looks forward to working together to build comprehensive tools for enterprises to safely and reliably implement AI applications. The completion of this acquisition is subject to customary closing conditions.