Anthropic has been selected by the UK’s Department for Science, Innovation and Technology (DSIT) to help build and pilot a dedicated AI-powered assistant for GOV.UK, marking a significant step in the UK government’s efforts to deploy advanced AI in public services.
The new assistant, powered by Anthropic’s Claude models, is designed to help citizens navigate government services and receive tailored, context-aware guidance. The initial focus will be on employment-related services, supporting people who are entering or re-entering the workforce by helping them find jobs, access training opportunities, and understand the support available based on their individual circumstances.
The project builds on a Memorandum of Understanding signed in February 2025 between Anthropic and the UK government, aimed at exploring how frontier AI can be applied safely and effectively to public-sector use cases. Since then, Anthropic and DSIT have been working closely to identify practical deployments that prioritize reliability, transparency, and user trust.
According to Anthropic, the GOV.UK AI assistant is an agentic system that goes beyond answering static questions. It will actively guide users through government processes, maintain context across interactions so users do not need to repeat information, and give individuals full control over what data is remembered. All personal data will be handled in line with UK data protection laws, with opt-out options available at any time.
We’re excited to partner with the UK government to help deliver on the AI Opportunities Action Plan,
said Pip White, Head of UK, Ireland and Northern Europe at Anthropic.
This partnership demonstrates how frontier AI can be deployed safely for the public benefit and sets a standard for how governments can integrate AI into essential services.
A key objective of the collaboration is also capability-building inside government. Anthropic engineers will work alongside civil servants and developers at the Government Digital Service, with the aim of ensuring the UK government can independently operate and maintain the system over time. The pilot follows DSIT’s “Scan, Pilot, Scale” framework, allowing for controlled testing and iteration before any wider rollout.
The initiative forms part of Anthropic’s broader investment in the UK, including collaboration with the UK AI Security Institute on model evaluation and safety, expansion of its London office, and partnerships with universities, enterprises, and public institutions. It also reflects a wider global trend of governments working with AI providers to deploy technology for public benefit, following similar Anthropic-backed education initiatives in Iceland and Rwanda.
Material by Veronika Atanasova
Source and image: Anthropic
