FOR IMMEDIATE RELEASE

UK regulators: deployers stay liable for what AI agents do; Securetron says accountability starts at the gateway

TORONTO, November 10, 2026 — The United Kingdom's four digital regulators have told organisations deploying AI agents that handing work to an autonomous system does not hand over the legal responsibility for what it does. Securetron, maker of the AI-FW governance gateway, says the statement points to a practical question every enterprise deploying agents now has to answer: where, exactly, is that responsibility enforced?

The joint statement, "Protecting people and businesses in an agentic economy", was published on 30 October 2026 by the Digital Regulation Cooperation Forum, whose members are the Competition and Markets Authority (CMA), the Financial Conduct Authority (FCA), the Information Commissioner's Office (ICO) and Ofcom. Its core line is direct: "An AI agent's autonomy does not diminish the responsibility of organisations that deploy it to comply with applicable law and relevant regulation."

The statement makes clear that existing frameworks already apply to agentic AI, including competition, consumer and data protection law, online safety, financial services, payments and cyber security requirements. It also warns that the regulators will act where they find breaches: "Where we identify conduct that we reasonably suspect breaches the law and risks harm to consumers or businesses, we will not hesitate to use our powers."

The CMA's accompanying consumer-law guidance for AI agents takes the same position in plainer terms: a business is responsible for what an AI agent does in the same way it is responsible for what an employee does, even if a third party designed or supplied the agent. Under UK competition and consumer law, fines of up to 10 percent of worldwide turnover can apply. The ICO has also opened a call for evidence on agentic AI, running from 8 October to 20 November 2026, to inform future guidance and a statutory code of practice on AI and automated decision-making.

"Regulators are not waiting for a new AI law to hold deployers to account," said Sarah Reid, Securetron. "The rules already exist, and the statement says so plainly. The gap is operational: an agent that can browse, buy, message customers and touch internal systems needs a control point where its identity is verified, its authority is limited, its traffic is inspected and every action is recorded. That control point is the gateway between the agent and the model it calls."

AI-FW sits between AI agents and the large language models they call. It authenticates agents with API keys, JWT, mTLS or Kerberos, enforces roles with RBAC, and connects SSO with OIDC and SCIM provisioning. Every prompt and response is scanned in real time for jailbreaks and prompt injection, PII, toxicity and data exfiltration, while semantic intent analysis scores requests against natural-language policies. Per-model routing, rolling risk scores with automatic blocking, and a metadata-only audit log give security teams a complete record of what each agent did, without storing raw content. The gateway is fail-closed by design and retains zero content.

AI-FW is available in a free Community Edition, free to use with a published allowance of 10,000,000,000 tokens per rolling 30 days, no licence key and no signup. The Community Edition is proprietary free-to-use software, not open source. Enterprise editions add compliance reporting, high-availability clustering and dedicated support.

About Securetron

Securetron (https://securetron.net) builds infrastructure security products for the AI era, including AI-FW, the self-hosted AI firewall and governance gateway for AI agents and LLM backends (https://aifw.io), PKI Trust Manager for enterprise certificate lifecycle management, and Boltron-AI, an AI agent for PKI operations automation.

Press contact:
Sarah Reid
Securetron
1 Yonge St, Toronto, ON M5E 1E6, Canada
+1 (438) 927-8036
[email protected]