OpenAI acquires Promptfoo to enhance AI safety testing

Investing.com – OpenAI announces plans to acquire Promptfoo, an AI security platform that helps companies identify and fix vulnerabilities in AI systems during development. The company states that once the acquisition is complete, Promptfoo’s technology will be integrated into the OpenAI Frontier platform, which is used to build and operate AI collaboration tools.

Promptfoo’s tools are currently used by over 25% of Fortune 500 companies. The platform includes an open-source command-line interface and a library for evaluating and red-teaming large language model applications. OpenAI says it will continue developing this open-source project within the enterprise features of Frontier.

The Promptfoo team is led by Ian Webster and Michael D’Angelo. According to Srinivas Narayanan, OpenAI’s Chief Technology Officer for B2B applications, Promptfoo brings engineering expertise in the assessment, security, and testing of enterprise-level AI systems.

OpenAI outlined several features planned for businesses using Frontier. The company will directly add automated security testing and red-teaming capabilities to the platform to help identify risks such as prompt injection, jailbreaking, data leaks, tool misuse, and policy violations. Security and assessment tools will be integrated into the development workflow, and the platform will also include reporting and traceability features to support governance, risk management, and compliance.

Promptfoo co-founder and CEO Ian Webster said the company’s goal is to provide developers with a practical way to protect AI systems. Webster stated that joining OpenAI will enable the team to accelerate its work in security, safety, and governance, supporting teams building AI systems.

This article was translated with the assistance of artificial intelligence. For more information, please see our Terms of Use.