OpenAI Acquires Promptfoo to Bolster AI Agent Security
Key Points
- OpenAI acquires Promptfoo, a security startup founded in 2024.
- Promptfoo’s tools protect large language models from adversarial attacks.
- The technology will be integrated into OpenAI Frontier, the enterprise AI agent platform.
- Founders Ian Webster and Michael D’Angelo built Promptfoo’s open‑source testing library.
- Promptfoo serves over 25% of Fortune 500 companies and has raised $23 million.
- OpenAI plans to use Promptfoo for automated red‑teaming, workflow security checks, and risk monitoring.
- OpenAI will continue supporting Promptfoo’s open‑source offerings.
- The deal highlights the growing focus on AI security as autonomous agents become more common.
OpenAI announced that it has acquired Promptfoo, a security startup founded in 2024 that protects large language models from adversarial attacks. The deal will integrate Promptfoo’s testing tools into OpenAI Frontier, the company’s enterprise platform for AI agents. Promptfoo, created by Ian Webster and Michael D’Angelo, already serves a significant share of Fortune 500 firms and has raised $23 million. OpenAI said the technology will enable automated red‑teaming, workflow security checks, and risk monitoring for its agentic products, while continuing to support Promptfoo’s open‑source offerings.
Acquisition Overview
OpenAI announced Monday that it has acquired Promptfoo, an artificial‑intelligence security startup founded in 2024. Promptfoo was created to protect large language models (LLMs) from online adversaries, offering tools that let companies test security vulnerabilities in LLM‑driven applications. The acquisition will see Promptfoo’s technology integrated into OpenAI Frontier, the company’s enterprise platform designed for autonomous AI agents that perform digital tasks.
Background on Promptfoo
Promptfoo was founded by Ian Webster and Michael D’Angelo. Since its inception, the company has built an open‑source interface and library that enable security testing for LLM‑based workflows. According to the source, Promptfoo’s products are used by more than 25% of Fortune 500 companies. The startup has raised $23 million in funding and was valued at $86 million after its most recent financing round in July 2025. OpenAI did not disclose the financial terms of the transaction.
Strategic Fit for OpenAI
The acquisition underscores OpenAI’s focus on ensuring that its frontier‑lab AI agents can be deployed safely in critical business operations. By adding Promptfoo’s capabilities, OpenAI Frontier will gain automated red‑teaming features, the ability to evaluate agentic workflows for security concerns, and continuous monitoring for risks and compliance needs. OpenAI also said it will continue to develop Promptfoo’s open‑source offering, signaling a commitment to community‑driven security solutions.
Implications for the AI Landscape
The move comes as the development of independent AI agents has generated excitement about productivity gains while also presenting new attack vectors for bad actors seeking to manipulate automated systems or access sensitive data. Integrating robust security testing directly into the agent platform aims to mitigate these risks and reassure enterprise customers about the safety of AI‑driven processes.
Industry Reaction
Industry observers note that the deal reflects a broader scramble among frontier labs to prove that their technology can be used safely at scale. The acquisition also highlights the growing importance of specialized AI security tools as more organizations embed LLMs into core operations.
Future Outlook
OpenAI’s statement indicates that Promptfoo’s technology will become a core component of Frontier’s security architecture, offering both automated and manual testing capabilities. The continued support for the open‑source version suggests that the broader developer community will benefit from advanced security tooling, potentially raising the overall security posture of LLM‑based applications across the industry.