The UK Parliament’s Business and Trade Committee used a 13 October evidence session to examine whether voluntary arrangements are sufficient for frontier AI. The official agenda brought together former laboratory researchers, policy representatives from Anthropic, OpenAI, Google DeepMind and Meta, and representatives of the AI Security Institute and government.

The committee framed the central questions clearly: when should a model be tested, how independent should that testing be, what dangerous capabilities should trigger intervention, and who should be able to delay or stop a release? Its briefing also highlighted a structural limitation in the current system. The AI Security Institute can test models when access is granted, but the committee said it may not compel disclosure or prevent deployment.

That distinction matters because voluntary access can produce valuable technical evidence without resolving the governance problem. A test conducted too late, on an incomplete model or without enforceable remediation may inform the public while leaving the release decision entirely with the developer.

The next policy step should be judged by the quality of the mechanism, not the strength of the rhetoric. Useful legislation would define thresholds, access rights, confidentiality protections, reporting duties and appeal routes. It would also separate independent evaluation from the commercial timetable of the company being evaluated.

Execution Level Governance- What audit-ready agent governance actually looks like
David Girvin, founder and CEO of Assury argues that model-in-the-loop review, AI governing AI, is fundamentally unreliable for regulated environments: even the best-performing models miss a meaningful share of violations, the reviewing model is typically provided by the same vendor being reviewed, and prompt injection or context poisoning can compromise both the acting agent and its supposed overseer simultaneously. He makes the case for deterministic, architecturally enforced controls instead, walking through Assury's approach of autonomy zones, session risk accumulation, and credential starvation, which lets a compromised agent be cut off from its tools instantly rather than relying on time-boxed access. The conversation touches on why David is sceptical of just-in-time credentialing as a solution for agent security more broadly, since agent sessions don't run on predictable human timescales, along with the current gap between how identity and security vendors are pitching agent protection and what he sees happening at the execution layer in practice. He also discusses the compliance and audit implications of probabilistic decision-making, arguing that regulated industries will increasingly need tamper-evident, hash-chained audit trails that can withstand scrutiny from auditors and regulators who are only beginning to understand agentic risk, and reflects on a named frontier lab's own published framework as an example of the gap between research and practitioner reality. Elsewhere, David reflects candidly on building a bootstrapped security company in an increasingly crowded market, why he turned down aggressive VC funding to stay in control of the product, and what a credible third-party assessment of his own gateway would need to look like given that Assury sits directly in the execution path for every customer's agents.
BrightTALK

Share this post
The link has been copied!