OpenAI Purchases AI Security Firm Promptfoo to Enhance Agent Safety

OpenAI revealed on Monday that it has completed the acquisition of Promptfoo, a cybersecurity company specializing in artificial intelligence protection that was established in 2024.

According to OpenAI’s announcement, Promptfoo’s security technologies will be incorporated into OpenAI Frontier, the company’s business-focused platform designed for AI agents, following the completion of the acquisition.

While autonomous AI agents capable of executing digital workflows have sparked enthusiasm about potential productivity improvements, they have simultaneously created new vulnerabilities that malicious actors could exploit to access confidential information or compromise automated processes. This acquisition highlights the urgent need for AI companies to demonstrate that their systems can operate securely in mission-critical business environments.

Promptfoo was established by founders Ian Webster and Michael D’Angelo with the mission of creating security assessment tools for large language models. Their offerings include both open-source interfaces and proprietary libraries that enable organizations to identify potential security weaknesses in their AI systems. The startup claims that over 25% of Fortune 500 corporations currently utilize their security solutions.

Financial data from PitchBook indicates that Promptfoo secured $23 million in funding since its inception and achieved an $86 million valuation following its latest funding round in July 2025. OpenAI has not revealed the financial terms of this acquisition.

The integration will enable OpenAI’s agent platform to conduct automated vulnerability testing, assess AI workflows for potential security issues, and provide continuous monitoring for risk management and regulatory compliance purposes. OpenAI has also committed to maintaining and expanding Promptfoo’s open-source initiatives moving forward.