evidoria

← Back to browse

Good practice Imported

The UK National Audit Office's 2024 Audit of Government AI — 37% of 87 Bodies Had Deployed AI, 21% Had a Strategy, and Only 7 Algorithm Records Were Published

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

Top 74% 47/100 · Ask Evidence Copilot about this practice

The NAO's March 2024 survey of 87 UK government bodies found 37% had deployed AI (74 use cases), 21% had an AI strategy, 70% cited skills as a barrier, and the ATRS register held 7 entries.

The UK National Audit Office's 2024 Audit of Government AI — 37% of 87 Bodies Had Deployed AI, 21% Had a Strategy, and Only 7 Algorithm Records Were Published

Details

Promoter
UK National Audit Office
Period
2024
Keywords
public audit, AI governance, algorithmic transparency

Description

In March 2024 the UK National Audit Office (HC 612) assessed how well the Cabinet Office and DSIT had set up government to capture AI's benefits and manage its risks. Its survey of 87 government bodies (autumn 2023) found 37% had deployed AI, typically one or two use cases each, 74 in total; a further 37% were piloting or planning, and 70% reported skills as a barrier.
Only 21% of bodies had an AI strategy (61% planned one). Among 32 bodies with deployed AI, 24 usually had a named accountable owner, but only 15 usually identified AI use cases at organisational level before deployment. The draft public-sector AI strategy lacked a named owner between DSIT and the Cabinet Office, an implementation plan and performance metrics.
On transparency, the Algorithmic Transparency Recording Standard was not widely used: 38% of respondents never complied with it and the repository held only 7 published entries against 74 deployed use cases (per a law-firm analysis of the report). DSIT announced in February 2024 that ATRS would become mandatory for departments.
The audit is a governance and readiness review, not an outcome evaluation: it did not measure service benefits. CDDO's indicative estimate that almost a third of routine civil-service tasks could be automated was not tested for feasibility or cost. Its value to other administrations is a replicable survey-based baseline and a checklist of governance gaps.

Read the full analysis: https://www.nao.org.uk/reports/use-of-artificial-intelligence-in-government/

Implementation

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

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