Patronus AI raises $50M for AI evaluation
Patronus AI raised a $50M Series B and launched Digital World Models for testing AI systems in simulated environments.
AI evaluation is becoming one of the most important parts of the enterprise AI stack.
What happened
Patronus AI raised a $50M Series B led by Greenfield Partners and unveiled Digital World Models, a simulation-based approach for testing AI systems.
The company focuses on helping teams evaluate model behaviour before AI systems are used in production workflows.
Why it matters
As companies deploy more AI agents and models, they need better ways to test performance, reliability and failure modes. Simple benchmark scores are not enough when AI systems are making decisions inside real business processes.
Simulation environments could give enterprises a safer way to understand how AI behaves before exposing it to customers, employees or sensitive workflows.
The bigger picture
The AI market is moving beyond model access into the tools around the model. Evaluation, monitoring, governance and simulation are becoming core infrastructure categories.
Patronus AI’s round shows that testing AI is becoming a fundable business in its own right, not just an internal research task.