White Circle Lands $11M to Monitor AI Systems

The big picture: Paris-based AI control platform White Circle has secured $11 million in seed funding from leading operators and investors across OpenAI, Anthropic, DeepMind, Hugging Face, Mistral, Datadog, and Sentry. The company was founded by engineer Denis Shilov after his 2024 universal jailbreak exposed safety flaws in major AI models.

Why it matters:

• AI acceleration: AI is advancing rapidly, with models making critical decisions in sectors like healthcare and finance, yet many companies lack real-time visibility into their deployed AI systems.
• Security imperative: Enterprises need robust solutions to detect and mitigate risks such as hallucinations, prompt injection attacks, and harmful content in production AI environments.
• Compliance and trust: Ensuring AI reliability, security, and compliance is crucial for maintaining trust and preventing misuse as AI adoption expands across industries.

How it works:

• Real-time monitoring: White Circle provides a unified API to monitor AI inputs and outputs in real time, identifying issues like model drift and abusive user behavior.
• Automated enforcement: The platform allows companies to create custom enforcement policies and automate moderation actions, including rate limiting or banning users.
• Continuous improvement: White Circle learns from labeled feedback to enhance performance over time, supports over 150 languages, and is SOC 2 Type I and II certified and HIPAA compliant.

The catch: White Circle operates in a competitive AI observability and governance market, facing established firms like Arthur AI and WhyLabs. Larger infrastructure and monitoring companies, including Datadog and Scale AI, are also expanding their AI monitoring capabilities, increasing market pressure.