OpenAI Acquires AI Security Startup Promptfoo to Fortify Enterprise Agent Platform

OpenAI has agreed to acquire Promptfoo, a two-year-old startup specializing in security testing for large language models and AI agents, in a move that underscores the growing primacy of safety infrastructure in the race to deploy autonomous AI systems at enterprise scale.

The acquisition will integrate Promptfoo's automated red-teaming and vulnerability detection capabilities directly into OpenAI's Frontier platform, which enterprises use to build and operate AI agents. Promptfoo's technology systematically tests AI systems against adversarial attacks including prompt injections, jailbreaks, and sensitive data leaks within existing development workflows.

"Promptfoo brings deep engineering expertise in evaluating, securing, and testing AI systems at enterprise scale," said Srinivas Narayanan, CTO of B2B Applications at OpenAI. "Their work helps businesses deploy secure and reliable AI applications, and we're excited to bring these capabilities directly into Frontier."

Founded in 2024 by Ian Webster and Michael D'Angelo, Promptfoo raised $23 million in venture funding and was valued at $86 million following an $18.4 million Series A round in July 2025, according to PitchBook data. The startup's tools are already used by more than 25 percent of Fortune 500 companies, providing OpenAI with validated technology that addresses key barriers to AI adoption including compliance and risk management.

Financial terms were not disclosed, and the transaction remains subject to customary closing conditions. OpenAI said it will continue to develop Promptfoo's open-source command-line interface and library for testing LLM applications, balancing community contributions with proprietary enterprise features.

(The acquisition comes as security concerns around AI agents have intensified, particularly for systems connected to sensitive corporate data and operational tools. Experts have flagged numerous risks associated with autonomous AI platforms, especially as vendors target enterprise customers with agent-based offerings.)

The deal positions OpenAI to compete more aggressively in the enterprise AI market, where safety and governance features increasingly differentiate offerings. Nvidia is reportedly preparing to launch an open-source AI agent platform called NemoClaw, according to Wired, and has begun pitching enterprise software companies including Salesforce, Cisco, Google, Adobe, and CrowdStrike on partnerships. That platform is expected to include security and privacy tools, reflecting industry-wide recognition that enterprise adoption depends on robust safeguards.

The Promptfoo acquisition also validates investor bets on AI security infrastructure and may accelerate merger activity in the testing and governance segment. SecurityWeek tracked 426 mergers and acquisitions in 2025, including 13 involving AI companies, up from eight deals in 2024. By embedding security testing directly into its enterprise platform, OpenAI is setting a potential benchmark that could compel rivals to bolster their own capabilities as organizations prioritize governance in AI workflows.