Canada Revenue Agency — AI-driven detection of unwarranted GST/HST refund claims
Canada
Canada Revenue Agency uses ML and AI-powered analytics to detect unwarranted GST/HST refund claims. Since 2017–18, over C$1.1 billion in …
United Kingdom · London · See the United Kingdom profile · See the London profile
Top 62% 53/100 · Ask Evidence Copilot about this practice
Introduced in 2010 and used by about 4,300 staff, Connect cross-matches financial data to flag evasion; an FOI response reports roughly £3.4bn average and £4.6bn additional yield in 2024-25.
Connect is HMRC's analytics system, introduced in 2010 and grown over 15 years into one of the largest datasets held by the UK government. It draws on data from a range of financial sources to analyse tax returns and flag potential evasion for compliance staff.
According to a Freedom of Information response reported by Pinsent Masons in October 2025, about 4,300 HMRC staff use Connect, it supported more than 500,000 cases in the latest year, and it generated about £4.6bn in additional tax in 2024-25 against an average of about £3.4bn a year. HMRC also confirmed a partnership with a US analytics firm intended to bring more advanced AI into its data work.
The evidence is thinner than the headline: the figures come from an FOI reply, not a published, audited HMRC report; it is not stated whether yield is gross or net or how it is attributed to Connect versus other methods; and the 2019 ministerial figure cited was about £1.4bn. Little is public about safeguards, error rates or taxpayer redress, so the case is best read as a large-scale, self-reported enforcement tool whose independent evaluation remains outstanding.
Read the full analysis: https://www.pinsentmasons.com/en-gb/out-law/news/hmrc-big-data-tax-revenue-boost-4-6-billion
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.
Where this practice's information was retrieved from, and when.
Canada
Canada Revenue Agency uses ML and AI-powered analytics to detect unwarranted GST/HST refund claims. Since 2017–18, over C$1.1 billion in …
Ireland
Since 2011, Ireland's Revenue Commissioners have used AI risk scoring (REAP) to target tax-compliance interventions; 290,000 data-analytics-driven interventions in 2023 …
Latvia
Latvia's SRS uses ESCORT ML to score every registered company for fraud risk across tax and labour domains. Fraud detection …
South Africa
South Africa's SARS uses ML models across 1.2 billion annual records to select audits and detect refund fraud. In FY2024/25 …
Open full copilot Grounded in cited practices — always check the sources.