# UK’s third-party AI assurance roadmap mapped professionalization, information access and a planned innovation fund The UK published a third-party AI assurance roadmap on September 3, 2025, setting out government actions and future-market infrastructure rather than a new audit certification. Canonical: https://rangoon.ai/insights/archive-uk-trusted-third-party-ai-assurance-roadmap-september-2025/ Author: Rangoon Editorial (https://rangoon.ai/insights/editorial/) Published: 2026-10-24 Updated: 2026-10-24 Event date: 2025-09-03 Tags: [UK government](https://rangoon.ai/insights/tags/uk-government/), [AI governance](https://rangoon.ai/insights/tags/ai-governance/), [AI safety](https://rangoon.ai/insights/tags/ai-safety/), [AI security](https://rangoon.ai/insights/tags/ai-security/) Source note: Historical reporting: DSIT published the Trusted third-party AI assurance roadmap on September 3, 2025. The roadmap describes government actions to support a third-party assurance market, including a future consortium, skills and competencies work, information-access mapping, and an £11 million AI Assurance Innovation Fund whose first application round was stated to open in spring 2026. It also discusses potential quality-assurance models and their limitations. The roadmap did not establish a new mandatory auditor license, certify an assurance provider, accredit a model, open the fund on the publication date, or demonstrate that third-party assessments are uniformly high quality. Image: Third-party assurance inputs. Original Rangoon explanatory diagram: Source boundary: dated primary release reporting; original editorial analysis. No vendor interface, photograph, measurement result, or endorsement implied. ## Key takeaways - The roadmap set out market-building actions and future options, not a new mandatory assurance certification. - DSIT planned a consortium, skills framework and information-access work to support more consistent third-party assurance. - The £11 million innovation fund was announced with a future spring-2026 application round, so it was not a September 2025 grant-delivery result. ## DSIT published a market-development roadmap rather than a finished assurance regime [source 1](https://www.gov.uk/government/publications/trusted-third-party-ai-assurance-roadmap) [source 2](https://www.gov.uk/government/publications/trusted-third-party-ai-assurance-roadmap/trusted-third-party-ai-assurance-roadmap) On September 3, 2025, the Department for Science, Innovation and Technology published its Trusted third-party AI assurance roadmap. The policy paper says it sets out ambitions for the UK third-party assurance market and immediate government actions to support that emerging sector. It identifies quality, skills, information access and innovation as barriers, then discusses interventions and potential quality-assurance models. This matters because “roadmap” language sets a different state from a binding regulatory scheme or completed accreditation system. The publication did not name a required auditor license, publish a mandatory model-certification process or declare that market participants had met an official quality threshold. The event was a public policy paper that mapped work government intended to do and questions it intended to explore in support of a developing assurance market. ## A planned consortium was assigned building blocks for a future profession [source 2](https://www.gov.uk/government/publications/trusted-third-party-ai-assurance-roadmap/trusted-third-party-ai-assurance-roadmap) The roadmap says DSIT would support establishment of a stakeholder consortium, led by a nominated partner and supported by government, to work toward an AI-assurance profession. Its first-year tasks were described as developing a voluntary professional code of ethics, a skills and competencies framework, and a map of information-access requirements. The paper says those building blocks could later support work toward a certification or registration scheme, with auditing a possible starting specialization. That sequence is deliberate: convening and defining expectations come before a future professional scheme. The source does not say the consortium had already formed, that the ethics code had been adopted, or that an assessor held a new certificate on September 3. It records planned professionalization activity, not a completed credentialing outcome. ## The roadmap treated access to system evidence as an assurance prerequisite [source 2](https://www.gov.uk/government/publications/trusted-third-party-ai-assurance-roadmap/trusted-third-party-ai-assurance-roadmap) The policy paper says third-party assurance providers need information about AI-system functionality and use, inputs and outputs, algorithms and parameters, oversight and change management, and surrounding governance documentation. It proposes mapping those needs with providers and using the work to inform best-practice guidance for firms procuring assurance services. This is an important public-sector-adjacent assurance point: an external conclusion cannot be stronger than the information, scope and access supporting it. Yet the document does not settle how each proprietary model must disclose data, set a universal evidence package, or resolve conflicts between access and security, privacy or intellectual-property constraints. It also does not provide the results of a specific third-party review. The September event identified an evidence-access problem and a proposed mapping response; implementation and practical sufficiency remained future work. ## The announced innovation fund had a stated future application round and no September grant results [source 1](https://www.gov.uk/government/publications/trusted-third-party-ai-assurance-roadmap) [source 2](https://www.gov.uk/government/publications/trusted-third-party-ai-assurance-roadmap/trusted-third-party-ai-assurance-roadmap) The roadmap says DSIT was establishing an AI Assurance Innovation Fund with £11 million to support novel mechanisms that help actors identify and address risks from high-capability AI systems. Crucially, it says the first round would open for applications in spring 2026. That makes the September 2025 fact an announcement of a fund and planned timeline, not proof that applications were open, grants were selected or assurance mechanisms had improved. The paper also notes that quality infrastructure was still developing and that relevant accreditation standards were limited in number and scope. Those caveats prevent a sweeping conclusion that the UK market had already solved quality or professionalization. The defensible historical headline is a roadmap with announced actions and future funding, not a completed assurance regime or verified audit market. ## Sources 1. [GOV.UK: Trusted third-party AI assurance roadmap](https://www.gov.uk/government/publications/trusted-third-party-ai-assurance-roadmap) 2. [GOV.UK: Trusted third-party AI assurance roadmap (HTML)](https://www.gov.uk/government/publications/trusted-third-party-ai-assurance-roadmap/trusted-third-party-ai-assurance-roadmap) ## Related reading - [UK published an AI cyber-security code with 13 voluntary principles and an implementation guide](https://rangoon.ai/insights/archive-uk-ai-cyber-security-code-january-2025/) - [The UK retitled its AI Safety Institute and put security risks at the centre of its remit](https://rangoon.ai/insights/archive-uk-ai-security-institute-february-2025/) - [OECD launched the Hiroshima AI reporting framework, making comparable company disclosures the event—not proof of compliance](https://rangoon.ai/insights/archive-oecd-hiroshima-ai-reporting-framework-february-2025/) ## More information - [Documentation index](https://rangoon.ai/llms.txt) - [AI access and policies](https://rangoon.ai/ai/) - [Source repository](https://github.com/hypler-dev/rangoon)