The department, which is understood to be exploring a range of possible use cases of generative AI tools, has recently trialled technology to understand ‘causes and effects’ of customer interactions
HM Revenue and Customs has concluded an experiment testing the possibilities of using artificial intelligence to better understand the “causes and effects” of citizens’ engagement with the department.
Newly published procurement information reveals that, over a three-month period beginning on 28 March, the department conducted a trial of the use of the decisonOS software from AI firm causaLens. According to the company’s website, the tool enables organisations to “build, deploy, and measure the effectiveness of your decision-making in the real world” and “identify the cause-and-effect relationships within your data through causal discovery”.
According to a procurement notice, HMRC intended to deploy the technology to undertake a “proof-of-concept using causal AI to analyse and explain causes and effects of different customer processes”.
Related content
- GOV.UK rolls out Google AI-powered semantic search
- Tony Blair: ‘AI could automate 40% of DWP tasks’
- ‘We’re not moving fast and breaking things’ – ChatGPT-powered GOV.UK chatbot trials show promise but uncover ‘issues of accuracy and reliability’
The text of the contract – which was valued at about £19,000 – outlines that the tax agency wished to test the potential of AI to further its understanding of a particular “use case of interest”.
However, having the completed the three-month experiment at the end of June, PublicTechnology understands HMRC has decided not to take forward the proof-of-concept into any further deployment. The department believes that generative AI has significant potential, though, and continues a range of potential use cases in its operations.
“Innovation is a core HMRC value,” a spokesperson for the department said. “We champion new and different ways of working, helping us to operate efficiently and effectively.”
Causal AI is described by causaLens as “the only technology that can reason and make choices like humans do [and] utilises causality to go beyond narrow machine learning predictions and can be directly integrated into human decision-making”.
