IBM and EY have launched EY.ai for Tax, a set of artificial intelligence (AI)-powered compliance tools developed with the former’s watsonx platform.
The release is part of the ongoing collaboration between the two companies, combining IBM’s AI technologies with EY’s tax domain knowledge to support enterprise compliance processes.
The EY.ai for Tax platform incorporates open-source AI models, including IBM Granite, to assist organisations in automating tax-related processes and managing global tax data.
According to IBM and EY, the solutions are intended to address challenges associated with resource constraints, fragmented enterprise resource planning (ERP) systems, and complex regulatory environments.
EY Americas tax deputy vice chair Martin Fiore said: “Our professionals are extending beyond world-class tax technical knowledge and combining their experience with emerging technologies that can produce highly effective outcomes for our clients.
“The success of the alliance has been proven with the deployment of EY.ai for tax, built with IBM watsonx, with the IBM tax department. We flipped the long-held view of tax as an administrative function to a value creation function for businesses with AI-enabled solutions that are highly cost effective and performant all while prioritizing governance and trust.”
Three initial tools are included in the platform. The Intelligent Tax Data Lake aggregates, refines, and consolidates transactional data from multiple sources to support tax filings.
Built using watsonx.data, watsonx.ai and open-source models, the tool has been used within IBM’s tax department to consolidate data from 36 sources. IBM stated that the tool’s integrated data controls and reconciliation features have enabled higher-quality data outputs by automating manual procedures.
The Detect and Correct with Business Documents tool extracts and compares information from unstructured invoices against ERP data to support tax determinations and filings. It uses watsonx.ai alongside Docling, an open-source toolkit from IBM Research.
IBM indicated that the solution increases the number of source documents reviewed compared to previous manual efforts.
The third tool, Withholding Tax Determinations, also developed using watsonx.ai, automates the monthly task of determining appropriate withholding tax rates for individual transactions. IBM reported that the tool enables faster and more accurate evaluations of thousands of intercompany transactions.
IBM’s tax department is currently using all three solutions. The company estimates that their deployment could save tens of thousands of hours annually by replacing manual processes.
IBM has stated that it is working towards automating 80% of its foreign tax compliance obligations. The tools are designed with features that prioritise data security and compliance to help manage sensitive tax information and reduce the risk of penalties.
The tax compliance platform is part of a broader strategic alliance between IBM and EY, aimed at building AI-driven enterprise solutions. The companies have indicated that these tools can be deployed in varied client environments to assist with complex data management and regulatory reporting requirements.
IBM vice president and chief tax officer Kanthi Morrissey said: “We manage massive volumes of granular data from disparate sources to file thousands of tax returns across hundreds of legal entities worldwide.
“By combining EY domain expertise in tax and IBM’s AI technology, we were able to transform the way our tax department handles global tax compliance, freeing up the team to focus on more strategic work.”
