Fundraise

Disseqt raises $8M to police rogue AI agents

What's the deal? Disseqt, a London-based startup building governance tools for AI agents, has raised $8 million in a round led by Gresham House VenturesDealroom has a profile for this one. Try Dealroom →. Former OpenAI and Meta executives also backed the company, according to Sifted.

What's the endgame? Disseqt operates what it calls an AI Assurance Layer that sits between a company's AI agent deployments and its risk function. The platform runs red-team testing before deployment, monitors agents in real time, and produces audit-ready outputs across the deployment lifecycle.

It targets regulated enterprises in finance, insurance, and healthcare. Commercial operations are run from Dublin, with an engineering team in Bangalore and customer-facing staff across the UK, US, Ireland, and India.

Why now? The round follows OpenAI's disclosure that its AI agents breached Hugging Face in July 2026. As companies rush to deploy autonomous agents, investors are backing tools that police what those agents can access and do.

What could go wrong? The market is still testing whether dedicated "guardian" systems can reliably control the agents they oversee. Disseqt's pitch depends on compliance and security teams trusting one layer of software to catch failures across many others.

The signal: Money is flowing into a new governance layer for enterprise AI. In September 2026, cybersecurity firm Island raised $400 million at a $6.4 billion valuation, citing agent-related risks — a sign that investors see policing AI agents as its own category.

Read more: Sifted

Image credit: "MISL lab work" by uwnews, licensed under CC BY-ND 2.0

More top stories