OpenAI to buy cybersecurity startup Promptfoo to better safeguard AI agents

In a significant move to enhance the security of artificial intelligence systems, OpenAI announced on Monday its acquisition of cybersecurity startup Promptfoo. This strategic decision aims to strengthen the protection and testing of sophisticated AI technologies. While the financial details of the acquisition remain undisclosed, Promptfoo's team is set to integrate into OpenAI's operations. Promptfoo specializes in developing security tools designed to safeguard AI agents, bringing essential capabilities to OpenAI's Frontier platform. Ian Webster, CEO of Promptfoo, emphasized the increasing importance of securing AI systems as they become more interconnected with real-world data. He stated that joining OpenAI allows for accelerated advancements in security, safety, and governance for teams engaged in developing practical AI solutions. Additionally, OpenAI plans to continue supporting Promptfoo's widely-used open-source project, which enables developers to test various AI prompts and evaluate the performance of major language models, including those from GPT, Anthropic's Claude, and Google's Gemini. Previously, in July, Promptfoo secured $18.4 million in a Series A funding round led by Insight Partners, with participation from Andreessen Horowitz. With a total funding of $22.68 million and a post-money valuation of $85.50 million as of July 2025, the startup consists of a tight-knit team of 11 employees. Andreessen Horowitz has been actively investing in infrastructure and defense sectors, recently announcing a $15 billion fund aimed at fostering 'American Dynamism,' with a significant portion allocated for growth and infrastructure projects.