This article covers White Circle, an AI startup, which has raised £8m in a growth funding round to accelerate development of a platform that monitors and controls deployed AI models in real time. The funding aims to scale AI observability and safety tools for organisations deploying generative models, including enterprise teams and regulated sectors.
White Circle, an AI startup, has raised £8 million in a growth funding round to accelerate development of a platform that monitors and controls deployed AI models in real time — a sign that investors and customers are prioritising observability and safety as AI moves into production.
As organisations increasingly rely on generative models for decisions in hiring, healthcare, finance and customer service, tools to detect misuse, bias and model drift have become critical. White Circle’s raise reflects demand for third‑party systems that sit between AI providers and enterprise applications to catch hallucinations, prompt‑injection attacks and other behavioural issues before they affect users or regulators.
The company’s origins also underline the market pull. Founder Denis Shilov built White Circle after a 2024 public demonstration showed how a single prompt could bypass safety filters in major models, prompting interest from model providers and enterprise teams worried about production risks.
White Circle’s platform runs through a single API and applies proprietary monitoring models to live AI inputs and outputs. Features described by the company include real‑time policy enforcement, detection of sensitive data leaks, blocking of malicious agent instructions, identification of model drift and flags for abusive users. The platform supports more than 150 languages and uses labelled feedback to iteratively improve detection and enforcement.
Customers named in the announcement include consumer brand Lovable and two unnamed digital banks, suggesting early traction both in consumer tech and regulated financial services. The business also publishes empirical research — notably CircleGuardBench in 2025 and a later KillBench study covering over one million experiments across 15 models from providers such as OpenAI, Google, Anthropic and xAI — aimed at measuring moderation and behavioural differences between models.
White Circle said it raised £8 million from a group of individual and institutional backers. Named participants include Romain Huet (OpenAI); Dirk Kingma (ex‑OpenAI, now Anthropic); Guillaume Lample (Mistral); Thomas Wolf (Hugging Face); Olivier Pomel (Datadog); François Chollet (Keras); Mehdi Ghissassi (ex‑DeepMind); Paige Bailey (DeepMind); and David Cramer (Sentry).
Investors cited the team’s technical credibility and the company’s empirical research as reasons for backing the product-market fit in AI safety and observability. The funding will be used to speed product development and expand the team across the US, UK and Europe while growing the customer base.
In the announcement, Ophelia Cai, Partner at Tiny VC, said:
Denis and the White Circle team have an unusual combination of deep technical credibility and a clear commercial instinct. The KillBench research alone shows what's possible when you approach AI safety empirically rather than ideologically and the team is building the infrastructure the industry genuinely needs.
If you're researching potential backers in this space:
White Circle was founded by Denis Shilov after his 2024 jailbreak demonstration attracted widespread attention and an invitation from Anthropic to join its bug bounty programme. The experience exposed enterprise demand for systems that can provide visibility and control over what models do once they are embedded in products.
In the announcement, Denis Shilov, Founder & CEO at White Circle, said:
AI is moving faster than our ability to guide it. We already trust it with decisions that touch millions of people – from hiring to healthcare, finance and security – and with the rise of vibe coding, anyone can ship an AI product without knowing what that model is actually doing once it's live. Until now there’s not been a platform purpose-built to monitor AI’s behaviour, catch it when it goes wrong, or shape how it acts. With White Circle, we’re finally giving companies everything they need to hold their AI accountable and optimise their models in a single place, without sacrificing security, compliance or risk.
Product design is explicitly aimed at non‑technical and technical teams alike. In the announcement, Elena Iumagulova, Head of Design at White Circle, said:
We wanted any team, technical or not, to be able to see exactly what their AI is doing in one place – whether they're getting visibility into their models for the first time or running AI at scale and need to catch problems fast. Everything from real-time flagging to analytics and optimisation lives in one place, connected via a single API.
White Circle’s research output and the investor line‑up reflect a broader shift: firms building production‑grade monitoring, governance and safety tooling are moving from niche to mainstream. The KillBench findings — which report systematic behavioural differences between models on attributes such as nationality, religion, body type and phone brand — add empirical weight to calls for robust testing and oversight.
Regulatory scrutiny in the UK and EU, combined with enterprise risk appetites, is pushing organisations to adopt third‑party observability systems rather than rely solely on model vendors. Tools that provide audit trails, policy enforcement and multilingual coverage are therefore becoming part of the standard stack for companies shipping AI.
This funding round, and White Circle’s hiring plans across the US, UK and Europe, illustrate how the market for AI safety and observability infrastructure is maturing and attracting both technical founders and experienced backers from major AI organisations.
Click here for a full list of 7,589+ startup investors in the UK