From automation to autonomy: How agentic AI will transform tax workloads
From automation to autonomy: How agentic AI will transform tax workloads
While agentic AI is still in its early stages, it won’t be long before it can routinely handle a variety of complex tax tasks and specialise in specific tax domains.
Today, artificial intelligence (AI) is being deployed by organisations operating in every industry sector to automate routine tasks, create content, mitigate fraud, summarise and perform analysis on large data sets. And the tax and accounting industry is no exception.
For example, the rapidly evolving field of generative AI (GenAI) has started to enable tax and finance teams to:
- automate tasks such as data entry and the generation of reports
- automate data gathering, cleansing and categorisation
- transform tax research, generating succinct and easy-to-read, detailed summaries.
Until now, the tax industry has been cautious about embracing AI for a variety of reasons, including concerns around data security, a lack of trust around AI outputs and worries that AI needs too much human oversight to be reliable. However, that is set to change with the emergence of agentic AI.
Agentic AI: The next frontier
Capable of operating independently, making decisions and performing tasks with minimal human intervention, agentic AI systems can learn from past experiences and adapt to new situations and environments.
Unlike previous task-specific iterations of AI and automation technologies that follow prescribed steps, agentic AI is an autonomous entity that can execute multi-step processes independently and solve complex problems.
So, whereas traditional automation might flag an issue, agentic AI uses sophisticated reasoning to act on this information. For example, when dealing with an indirect tax issue, it can identify when an expenses claim featuring an overnight stay lacks key information and automatically email the accommodation provider using local language and descriptors to request a duplicate document containing the required invoice number or VAT details.
While agentic AI is still in its early stages, it won’t be long before it will be routinely handling a variety of discrete tasks. These use cases include the following.
Streamlining tax compliance – autonomously tracking and monitoring changes in tax law to ensure filings meet regulatory standards.
Preparing and reviewing tax computations and undertaking financial auditing – flagging and resolving discrepancies, creating and executing checklists and preparing internal audit reports. Saving time and effort, agentic AI then serves up review information for a human to assess.
Researching complex tax issues – analysing financial data and preparing initial reports. Utilising these insights, tax and accounting professionals will be able to dedicate more time to providing value-add strategic advice to internal or external stakeholders.
Automated tax planning – identifying tax-saving opportunities and drafting documentation to support next steps.
Finance audits – autonomously identifying missing entries, resolving discrepancies, reducing time to closure and generating audit-ready reports.
Workflow optimisation – managing tax deadlines, coordinating teams and integrating actions across multiple platforms to streamline the initiation and completion of filings.
AI agents can also become specialists in specific tax domains, such as VAT, transfer pricing or international compliance. Acting like digital team members, agents will be able to collaborate and refer complex issues to the “specialist” agents. By sharing insights and updates, they collectively evolve, forming a high-performing, always-learning tax workforce.
Integrated into specialised tax and accounting tools and solutions, agentic AI is already complementing existing AI capabilities, enabling organisations to automate workflows, enhance audit quality and delivery, and monitor regulatory changes in real time. Poised to reshape how tax and accounting firms undertake work, there is little doubt that agentic AI will enable organisations to operationalise AI at scale and achieve measurable outcomes.
This article is sourced from the following link:
