Ex-Palantir and AI researchers raise $7.3M for HelmGuard to automate enterprise risk
Tech Funding News Sofia Chesnokova ● Covered by 2 sources
HelmGuard raised $7.3M to use AI agents for enterprise compliance. It’s betting checklists are outdated, and even one of Vanta’s backers signed up.
Based on reporting by Tech Funding News, Sofia Chesnokova — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
HelmGuard has pulled in $7.3 million in seed funding to build an AI-agent platform for governance, risk and compliance. The pitch is blunt: stop treating compliance like paperwork that humans shuffle around, and let agents do the checking themselves.
The company was started by John Daley and Jack Miller, who met through an Entrepreneur First programme organised by Matt Clifford. Daley spent eight years at Palantir after training as a technology lawyer; Miller comes from AI research and worked on some of the early large language models at OpenAI and Google while still a student. Clifford later became HelmGuard’s first investor, backing the pair’s “defensive acceleration” view that better defensive tech is the right answer to AI risk.
HelmGuard is taking aim at established names such as Vanta, Drata and Secureframe, but it says the difference is deeper than a feature list. Daley describes the company’s internal line as “decisions over documents”: instead of creating checklists for people, HelmGuard wants AI agents to act on them. Miller argues that older compliance platforms have accumulated so many modules and add-ons that they’ve become too fragmented for agents to use cleanly.
The startup says it ties everything into a single graph covering internal compliance, third-party risk, AI governance, agents, employees and business units. To keep the agents from wandering off the rails, Miller points to three guardrails: a confidence score with each assessment, citations for each conclusion, and human review when a decision is critical. Customers can set the system so that a person only steps in once risk crosses a chosen level.
HelmGuard says that approach has already scaled in practice. One US insurance client used it to assess 1,250 counterparties in under a week, and London-based Callosum used it to set up its security and compliance programme from scratch. The company is based in King’s Cross, London, plans to open a US office soon, and expects to raise a Series A in 12 to 18 months.
My take — AI-written commentary, not fact-checked reporting
This is the right fight: compliance software has spent years producing prettier paperwork and calling it progress. The awkward bit for the old guard is that agents are a much better fit for the job than endless dashboards and annual-certification theatre. If risk is happening at runtime, the tools should be too.
Read more about this at: Tech Funding News