White Circle, a company focused on making AI safer and more controllable in production, has raised $11 million in seed funding. The startup aims to help organizations manage AI models more securely and effectively as AI deployments grow rapidly across industries. The new funding will support product development and hiring in the US, UK, and Europe.

Key Backers and Industry Experts

The funding round includes investments from senior figures at leading AI companies such as OpenAI, Anthropic, DeepMind, Hugging Face, Mistral, Datadog, and Sentry. Notable individuals like Romain Huet, Guillaume Lample, Thomas Wolf, and François Chollet are among the backers. These experts have experience building and policing AI models, which aligns with White Circle’s mission.

Many of the investors have directly worked on or with the AI models that White Circle’s platform is designed to oversee. This strong backing from industry veterans highlights the importance of security and accountability in AI development today.

The Company’s Origin and Technology

White Circle was founded by engineer Denis Shilov, who gained attention in 2024 after creating a prompt that bypassed major AI safety filters. This demonstration went viral, leading to contacts from top AI labs and Shilov joining Anthropic’s bug bounty program. The platform was built to address the vulnerabilities his viral prompt revealed.

The platform acts as a single API layer that monitors AI inputs and outputs in real time. It can detect harmful content, hallucinations, prompt injections, model drift, and abusive users. Customers can customize enforcement actions like rate limits or bans, and they can use user feedback to improve accuracy. The system supports over 150 languages and complies with industry standards for security and privacy.

Shilov emphasizes that the platform helps fill an accountability gap that has widened as AI is deployed more widely. He says, “AI is moving faster than our ability to guide it,” and White Circle provides the tools companies need to ensure responsible AI use without sacrificing security or compliance.

Research and Industry Impact

Alongside its product, White Circle has published two research papers. One, called CircleGuardBench, tests how AI moderation models perform in real-world scenarios. The other, KillBench, ran over a million experiments across 15 AI models from major labs, revealing biases related to nationality, religion, and even phone brands in decision-making processes.

KillBench also found that structured-output integrations, common in production deployments, can reduce refusal rates but sometimes amplify biases. Industry observer and venture capitalist Ophelia Cai praised White Circle for building infrastructure that the AI industry needs, citing the team’s technical credibility and commercial instinct.

To date, White Circle has served more than one billion API requests. Its clients include a digital banking giant and a company named Lovable, although specific details are not disclosed. The company has not announced its valuation or revenue figures yet.

Overall, White Circle aims to set a new standard for AI safety and control at a time when AI’s influence continues to expand into critical areas of society and business. With strong backing and innovative technology, it’s positioned to shape the future of responsible AI management.

Inspired by

• https://thenextweb.com/news/white-circle-11m-seed-ai-control-platform

Sources

• whitecircle.ai
• huggingface.co
• whitecircle.ai