Bloomberg Tax and Accounting released two new generative AI-driven features to aid users in tax research.
The first is Bloomberg Tax Answers, which enables users to ask tax questions that the bot can then answer with supporting primary sources and expert analysis so accountants can find, validate and apply the information to their workflows. It provides what Bloomberg said would be a brief but meaningful answer to a user's search directly on top of regular search results, with no need to learn a new tool. Each answer generated includes citations and links to the Bloomberg Tax authorities and source documents used to generate it, including select primary and secondary sources such as the Internal Revenue Code, federal and state tax agency documents, state tax statutes and regulations, and Bloomberg Tax content.
The second is AI Assistant, a chat-based research tool that lets users ask specific questions from within a document, including Bloomberg Tax's own portfolios, as well as create data visualizations that can compare tax information across jurisdictions. Users can also ask specific questions about the document to quickly identify the information they are looking for.
"The latest AI-powered features for Bloomberg Tax & Accounting showcase our dedication to innovation and solving complex tax research challenges," said Evan Croen, head of Bloomberg Tax. "Bloomberg Tax Answers and AI Assistant deliver rapid, accurate answers and facilitate cross-jurisdictional comparisons. Additionally, users can verify information with direct access to cited source documents, enhancing reliability and trustworthiness."
Both solutions were initially released for select users in the new Bloomberg Tax Innovation Studio late last year (
Bloomberg Tax Answers and AI Assistant will be subject to ongoing refinement based on customer feedback. AI Assistant will be updated with additional research skills in the coming months. Both Bloomberg Tax Answers and AI Assistant are available within the Bloomberg Tax platform at no additional charge.