Promptfoo specializes in tools that allow developers to test, validate, and secure complex AI systems before they are deployed.
Promptfoo specializes in tools that allow developers to test, validate, and secure complex AI systems before they are deployed.
As AI agents increasingly interact with enterprise software, data systems, and automated workflows, ensuring that those systems behave safely and reliably has become one of the industry’s most pressing challenges. Promptfoo CEO Ian Webster said joining OpenAI will allow the company to accelerate work on “stronger security, safety, and governance capabilities” for teams building AI-powered applications.
OpenAI said it will continue supporting Promptfoo’s open-source testing framework, which enables developers to evaluate prompts, compare large language models, and measure the performance of systems like ChatGPT, Anthropic’s Claude, and Google’s Gemini. Those tools have become widely used within the developer community as organizations experiment with integrating AI into production systems.
The acquisition reflects a broader strategy by OpenAI to expand both its capabilities and talent base as competition intensifies across the AI sector. In recent months the company has been steadily acquiring startups and hiring prominent AI engineers, including bringing on Peter Steinberger, creator of the OpenClaw framework used to build autonomous AI agents.
OpenAI also acquired health-tech startup Torch earlier this year and purchased the AI interface startup Software Applications in late 2025.
Promptfoo raised an $18.4 million Series A round last July led by Insight Partners, with participation from Andreessen Horowitz. The startup has a small team of roughly 11 employees and was valued at about $85 million as of its last funding round.
The deal highlights a growing realization across the AI industry: as models become more powerful and agents begin to act autonomously, security, testing, and governance infrastructure will be just as important as the models themselves.
By bringing Promptfoo’s technology inside its platform, OpenAI is positioning itself not only as a builder of AI systems, but also as a provider of the safety frameworks required to deploy them at scale.
Jay Goldberg is a staff writer at TechEchelon covering technology, markets, and policy. He files the breaking news and deal coverage that move the publication's core desks.
More from Jay →CISA has confirmed active exploitation of CVE-2026-20253, a critical Splunk Enterprise flaw that allows unauthenticated remote attackers to manipulate files, and ordered federal civilian agencies to patch by Sunday under Binding Operational Directive 26-04.
Market intelligence platform Klue has confirmed attackers stole OAuth tokens connected to customer Salesforce environments on June 12, as the newly emerged Icarus extortion group publicly claims responsibility and the list of affected organizations continues to grow.
Human Rights Watch obtained Bulgarian export licensing records showing surveillance firm Circles sold phone-interception and geolocation tools to ten governments — including the UAE, Bahrain, and Azerbaijan — between 2018 and 2023, raising fresh questions about the EU's export control enforcement.
Written for readers who already know the basics — markets, AI, and the policy decisions that shape both.