How HM Revenue and Customs uses AI

HMRC's Debt Management function runs a machine learning model called Business Propensity to Pay that scores about 4 million business debtors each month on how likely they are to engage with debt recovery. A gradient boosted decision tree, built in SAS Viya, learns from debt value and past repayment behavior across VAT, PAYE and Corporation Tax debts linked at customer level.

The segmented score helps decide which collection journey a debtor receives, including the timing, frequency and content of the first letters. HMRC says the model replaced a simple segmentation based on exposure to the pandemic (before that segmentation, most customers received the same journey) and that the score is only one factor in the journey. The record discloses no recovery results.

HMRC staff use the Organisational Listening Tool to find themes in free text customer feedback. A supervised text classifier and BERTopic clustering group comments into topics, and a large language model writes labels for each topic from representative words and comments. Unsupervised models are reviewed and refreshed every three months.

HMRC's VAT Return Analysis Tool brings a VAT trader's entity, ledger and return data for the most recent seven years into one interactive view for VAT officers, and uses a classical statistical model (seasonal trend decomposition with an interquartile range rule) to flag anomalous values in the return history. Officers use it to prepare and carry out compliance checks; the tool makes no decisions, and any assessment is made by an officer and can be appealed through the normal route. HMRC's transparency record says around 5,500 officers are licensed and the tool is used about 1,500 times a day.

HMRC's digital assistant answers tax questions typed in plain language, matching them to intents with natural language understanding and replying with non personalized answers that link to GOV.UK guidance. It covers 60 of HMRC's more than 120 taxes, asks the user to choose between likely meanings when unsure, and needs no login. Complex questions escalate to webchat with a human adviser, who sees the whole prior conversation and only continues after identity and verification checks.

HM Revenue and Customs (HMRC), the UK's tax authority, receives thousands of handwritten and paper tax forms a day, such as agent authorizations, refund claims and tax credit transfers, that took over an hour per document to review manually and drove processing times of up to 45 days. HMRC deployed Hyperscience's intelligent document processing platform, on premises to meet UK government standards, to extract data from the forms with machine learning models trained on them and route only low confidence fields to a person. It reports that most forms are now processed in a day or less, and that it redeployed more than 300 staff to higher value work.