evidoria

← Back to browse

Good practice Imported

GOV.UK Chat — the UK Government's Generative-AI Assistant for Citizen Guidance

United Kingdom · London · See the United Kingdom profile · See the London profile

Evidence: Observational / pre–post Top 6% 87/100 · Ask Evidence Copilot about this practice

The UK's Government Digital Service piloted GOV.UK Chat, a retrieval-augmented generative-AI assistant answering citizens' tax, benefits and visa questions — accuracy rose from 76% to 90% across two public pilots totalling over 10,000 users and 26,000 questions.

76% → 90%
Measured accuracy (first benchmark to all-topics benchmark)
>10,000
Total users across two pilots
~26,000
Total questions asked
73%
Users finding responses useful
64%
User satisfaction
10.7 seconds
Average response time
508
Jailbreak attempts blocked
GOV.UK Chat — the UK Government's Generative-AI Assistant for Citizen Guidance

Details

Maturity
Scaling
Promoter
Government Digital Service (GDS), UK Department for Science, Innovation and Technology
Period
July 2023 – ongoing (national rollout continuing into 2026)
Keywords
AI chatbot, generative AI, digital government, citizen engagement, retrieval-augmented generation

Context

GOV.UK Chat is a retrieval-augmented-generation assistant built by the UK Government Digital Service (GDS) from July 2023 to answer citizens' tax, benefits and visa questions using only published GOV.UK guidance.

Results

Across two public pilots (over 10,000 users, ~26,000 questions), measured accuracy rose from 76% to 90%; 73% found responses useful and 64% were satisfied; average response time was 10.7 seconds; safety guardrails blocked 508 jailbreak attempts.

Implementation

Implementation detail (cost, timeline, staffing, conditions for success) is not yet available for this practice.

Commonly funded by

National / regional programmes

Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Attachments

Similar practices you may find useful