by Sam Swansborough
Business Tax Director
30 June 2025
generalby Sam Swansborough
Business Tax Director
The growing use of AI and automation in the UK tax system is set to reshape how businesses engage with compliance, reporting and reliefs. That was the clear message from this year’s Tax Technology Conference, hosted by the Chartered Institute of Taxation (CIOT) and Association of Tax Technicians (ATT), which brought together senior HMRC officials, professional firms and technology experts. With updates on Making Tax Digital, real-world use cases of AI by tax authorities, and warnings about ethical risks and digital readiness, the event laid out a roadmap that will be particularly relevant to SMEs, innovative businesses, and those navigating complex reliefs such as R&D tax credits.
Call for Simpler Tax Rules in a Digital Age
The keynote was delivered by Professor Michael Mainelli, formerly Lord Mayor of the City of London, who argued that the UK’s current tax and regulatory system is too complex to support effective automation. His advice to policymakers was succinct: simplify, then automate, then integrate. Mainelli compared the UK unfavourably with smaller economies like Ireland, Singapore and Switzerland, which have used simpler systems to attract global investment. He suggested that high tax complexity not only discourages inward investment but hampers domestic productivity, and that businesses should expect increasing scrutiny not just via the monitoring of rates and reliefs, but in the design and consistency of their compliance and reporting.
HMRC Sets Out Timeline for Digital Tax Reforms
Senior figures from HMRC used the conference to highlight the department’s accelerating shift towards a digital-first tax system. More than 70 percent of processing functions are now digitised, and the HMRC app, already used by over six million individuals, continues to receive some of the highest satisfaction scores among government services. But attention is now turning to income tax. From April 2026, self-employed individuals and landlords earning over £50,000 will be required to maintain digital records and file quarterly updates. This represents a major expansion of the Making Tax Digital programme, which has so far focused on VAT. Officials confirmed that a wider transformation roadmap will be published later this summer. The document is expected to set out long-term plans for corporation tax, Companies House filings, and further integration of software platforms, developments that could have far-reaching implications for SMEs. While concerns around digital exclusion were acknowledged, HMRC’s message was clear: the future of tax compliance in the UK will be real-time, data-led and increasingly automated.
AI in Tax Enforcement: Benefits and Boundaries
In the main panel discussion of the day, HMRC and external experts discussed how AI is already being used to improve compliance. AI models are analysing taxpayer data, flagging anomalies, and helping segment cases for follow-up. These tools are intended to support enforcement by detecting patterns that may be missed through traditional review methods. However, ethical concerns were a consistent theme. Speakers highlighted the risk of bias, the need for human oversight, and the importance of transparency in AI decision-making, particularly when enforcement action may follow. The panel highlighted that all AI tools in use are subject to legal safeguards, including Article 22 of the GDPR, which requires a human to be involved in automated decision-making. There was broad consensus that AI can enhance efficiency but must be deployed with care. Tax authorities may be able to process more data faster, but without appropriate controls, there is a risk that speed could come at the cost of fairness and clarity.
AI Tools Move From Theory to Practice in Business Settings
Beyond government, attention turned to how businesses themselves are beginning to adopt AI in everyday operations. Presenters from the private sector outlined a growing list of practical applications, from producing first-draft reports to summarising research and classifying financial transactions. Tools such as ChatGPT and Perplexity were cited as entry points for businesses looking to experiment with automation without the need for heavy investment in infrastructure. It was noted that while these tools can deliver efficiency gains, they require a strong understanding of their limitations. Conference contributors emphasised that businesses should avoid uploading sensitive data into public tools and should treat all AI-generated output as provisional, subject to internal review and human judgment. Internal policies governing the use of such tools are increasingly seen as a necessary safeguard. The UK’s AI Management Essentials framework, presented during a breakout session, may soon become a standard reference point for businesses adopting AI. With regulators preparing new guidance on responsible use, the shift from novelty to operational necessity appears to be well underway.
Perspective from the Profession
Speaking after the event, Sam Swansborough, Director in our Business Tax team, reflected on the practical implications for clients. He noted that while AI and automation may improve efficiency, they also raise the bar for accuracy and audit-readiness. “We’re not just seeing new tools, but a shift in how tax assessment works,” he said. “The system is becoming more real-time, more data-driven and more vigilant. For businesses preparing for MTD or their continued compliance, the quality of processes and documentation will matter more than ever.”
Sam added that the fundamentals of good tax governance remain unchanged, but expectations around transparency and consistency are only set to increase. “Technology can help but it doesn’t replace sound advice, clear records, or a proper review.”
Preparing for What’s Next
The conference made clear that the digitalisation of tax is no longer a future aspiration, it is happening now. Businesses relying on manual processes, spreadsheets or paper-based records should begin reviewing their systems to ensure they are compatible with the expectations of a digital-first tax authority. Those already using cloud software or engaging with reliefs such as R&D tax credits should take time to test whether their current processes are audit-ready in a data-led environment. Evidence, clarity and consistency are all likely to come under closer scrutiny as AI tools make it easier to identify gaps or inconsistencies. James Cowper Kreston is supporting clients through this transition, from planning for Making Tax Digital to strengthening the compliance and documentation processes that underpin reliefs and disclosures. As tax becomes more digital, staying ahead of the curve will not just reduce risk, it may also be a source of competitive advantage.