OpenAI integrates safety into AI agents by acquiring the startup Promptfoo

OpenAI acquires Promptfoo startup

OpenAI has entered into an agreement to purchase the two-year-old startup Promptfoo, which develops tools for testing AI safety early in its development. According to Bloomberg, the deal aims to reduce risks for clients using AI solutions and attract a large corporate market.

Who is Promptfoo?

- Founded: 2022
- Location: San Francisco.
- Services: open-source code for neural network testing, including “red‑team” where specialists simulate attacks on their own models, uncovering weaknesses before attackers do.
- Clients: about 25 % of companies in the Fortune 500 list.

What will change after integration?

Promptfoo’s technologies will be integrated into OpenAI’s corporate platform – Frontier. This will give Frontier automated testing, change monitoring, and a reporting system that meets security requirements.

Investor reaction

Ganes Bell (CEO Insight Partners), who led Promptfoo’s $18.4 million funding round last year, said: “The purchase demonstrates a clear commitment by OpenAI to building enterprise AI that is not only powerful but also safe and reliable at scale.” Financial details of the deal are not yet disclosed.

Future of Promptfoo

OpenAI will continue developing Promptfoo’s open source code alongside its commercial integration. The company has recently unveiled an AI agent capable of finding and fixing vulnerabilities in databases – a tool similar to Anthropic PBC.