{"id":102220,"date":"2025-01-30T08:02:48","date_gmt":"2025-01-30T16:02:48","guid":{"rendered":"https:\/\/xira.com\/p\/2025\/01\/30\/bloomberg-law-rolls-out-two-ai-powered-legal-research-tools-emphasizing-accuracy-and-source-attribution\/"},"modified":"2025-01-30T08:02:48","modified_gmt":"2025-01-30T16:02:48","slug":"bloomberg-law-rolls-out-two-ai-powered-legal-research-tools-emphasizing-accuracy-and-source-attribution","status":"publish","type":"post","link":"https:\/\/xira.com\/p\/2025\/01\/30\/bloomberg-law-rolls-out-two-ai-powered-legal-research-tools-emphasizing-accuracy-and-source-attribution\/","title":{"rendered":"Bloomberg Law Rolls Out Two AI-Powered Legal Research Tools, Emphasizing Accuracy and Source Attribution"},"content":{"rendered":"<p>Bloomberg Law has launched two new generative AI-powered research tools \u2013 Bloomberg Law Answers and Bloomberg Law AI Assistant \u2013 marking the company\u2019s most significant foray into providing its customers with AI-enhanced legal research capabilities. While Bloomberg Law had previously rolled out targeted gen AI applications, such as AI-powered complaint summaries on dockets, these new [\u2026]<\/p>\n<p><a href=\"https:\/\/www.bloomberglaw.com\/\" rel=\"nofollow noopener\" target=\"_blank\">Bloomberg Law<\/a> has launched two new generative AI-powered research tools \u2013 Bloomberg Law Answers and Bloomberg Law AI Assistant \u2013 marking the company\u2019s most significant foray into providing its customers with AI-enhanced legal research capabilities.<\/p>\n<p>While Bloomberg Law had previously rolled out targeted gen AI applications, such as <a href=\"https:\/\/pro.bloomberglaw.com\/insights\/company-news\/bloomberg-law-introduces-ai-powered-complaint-summaries-on-dockets\/\" rel=\"nofollow noopener\" target=\"_blank\">AI-powered complaint summaries on dockets<\/a>, these new tools are its first to deploy gen AI broadly within its platform for legal research and document analysis.<\/p>\n<p>The new features, released in beta after a year of testing through Bloomberg Law\u2019s Innovation Studio, aim to streamline legal research workflows by providing direct answers to legal questions and enabling document-specific querying. Both tools are available to current Bloomberg Law subscribers at no additional cost.<\/p>\n<p>\u201cWe took our time to bring these products to market,\u201d <a href=\"https:\/\/www.linkedin.com\/in\/bobbypuglia\/\" rel=\"nofollow noopener\" target=\"_blank\">Bobby Puglia<\/a>, chief product officer at Bloomberg Industry Group, told me in an interview. \u201cWe were very deliberate and patient in bringing these things to beta until we felt like we\u2019d had extensive user testing and that we\u2019ve got it to a place where we could confidently say that we could trust the answers.\u201d<\/p>\n<p>For users, the Bloomberg Law Answers feature will now appear at the top of their search results, but only if they opt in to activating the feature. The feature provides brief but detailed responses to legal queries, with extensive citation support and specific attribution for different components of each answer, Puglia told me.<\/p>\n<p><em>(I have not yet seen a demonstration of either feature.)<\/em><\/p>\n<p>The AI Assistant, meanwhile, functions as a document-specific tool that appears in a side panel when viewing primary or secondary sources. It allows users to generate summaries and ask targeted questions about the specific source documents they are viewing, with responses confined to information contained within that document.<\/p>\n<p>A key feature of both tools is their approach to source attribution, Puglia said. Unlike some competing products that may list sources only at the end of responses, Bloomberg Law has implemented what Puglia describes as discrete footnotes within the answer to give specific extracts of where this piece of the answer is coming from.<\/p>\n<h3>A Measured Approach<\/h3>\n<p>In developing these gen AI features, Bloomberg Law took a measured approach focused on accuracy and transparency, Puglia told me. The development process relied heavily on Bloomberg Law\u2019s legal data analysts \u2013 lawyers with practice experience and subject matter expertise who helped establish quality benchmarks for the system\u2019s responses.<\/p>\n<p><a href=\"https:\/\/i0.wp.com\/www.lawnext.com\/wp-content\/uploads\/2025\/01\/ai-assistant-screengrab.jpg?ssl=1\" rel=\"nofollow noopener\" target=\"_blank\"><img data-recalc-dims=\"1\" loading=\"lazy\" decoding=\"async\" class=\"alignright wp-image-48712\" src=\"https:\/\/i0.wp.com\/www.lawnext.com\/wp-content\/uploads\/2025\/01\/ai-assistant-screengrab-1024x1024.jpg?resize=500%2C500&#038;ssl=1\" alt=\"\" width=\"500\" height=\"500\" title=\"\"><\/a>He said the product uses both prompt shields and guardrails, with prompt shields ensuring that users\u2019 queries remain within appropriate legal contexts, and guardrails ensuring that the outputs are properly sourced from Bloomberg Law\u2019s content.<\/p>\n<p>\u201cWe will only return answers that are based on the content that we have within our system,\u201d Puglia said. \u201cWe do not go out and give general answers from outside of our sources.\u201d<\/p>\n<p>Bloomberg Law Answers draws its answers from both primary sources, including court opinions, statutes, and regulations, and secondary sources, such as practical guidance materials and books. Bloomberg Law is gradually expanding the content sets available to the AI tools, with each new addition undergoing extensive testing to maintain answer quality and reliability, Puglia said.<\/p>\n<p>Asked about the risk of hallucinations in the answers the AI is providing, Puglia acknowledged that hallucination risks remain inherent to gen AI technology, but said Bloomberg Law has focused heavily on managing those risks through both its content-grounding approach and extensive guardrails. With its prompt shields, the AI will explicitly decline to answer questions outside its scope.<\/p>\n<p>The tools use multiple AI models, Puglia said, including those from OpenAI and Anthropic, with Bloomberg Law selecting specific models based on particular use cases and performance requirements. The company maintains a team of machine learning engineers who evaluate new models as they become available.<\/p>\n<h3>Future Plans<\/h3>\n<p>In the coming months, Bloomberg Law plans to release an expanded version of its AI assistant that will enable queries beyond document-specific interactions to broader research capabilities. This feature will incorporate tools such as jurisdictional comparisons, chart building, and citation treatments, according to Puglia.<\/p>\n<p>Even longer term, Bloomberg Law\u2019s development team is exploring how these AI tools might reshape legal research workflows more fundamentally.<\/p>\n<p>\u201cIf you look longer term, the way the technology is evolving, I do think that the way that you approach research does have the potential to dramatically change from a workflow perspective,\u201d Puglia said. \u201d \u2026 We could more seamlessly integrate research into their day-to-day tasks, and it\u2019s not a separate standalone task.\u201d<\/p>\n<p>Bloomberg Law is also considering how its content and technology might integrate with other platforms and tools that law firms are using, including Microsoft\u2019s Copilot, recognizing that the traditional model of accessing legal research through a dedicated website may evolve, he said.<\/p>\n<p>I asked Puglia whether Bloomberg is providing training and support to users to help them transition into these new tools. He emphasized that user adoption of these tools is completely optional, and said that Bloomberg Law is working with firms to accommodate their specific compliance requirements and comfort levels with the technology.<\/p>\n<p>\u201cThe first thing we do is we don\u2019t force it,\u201d Puglia said. \u201cWe work with our clients to make sure that every firm still has their legal processes and reviews that they have to go through.\u201d<\/p>\n","protected":false},"excerpt":{"rendered":"<p>Bloomberg Law has launched two new generative AI-powered research tools \u2013 Bloomberg Law Answers and Bloomberg Law AI Assistant \u2013 marking the company\u2019s most significant foray into providing its customers with AI-enhanced legal research capabilities. While Bloomberg Law had previously rolled out targeted gen AI applications, such as AI-powered complaint summaries on dockets, these new [&hellip;]<\/p>\n","protected":false},"author":3,"featured_media":102221,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"_et_pb_use_builder":"","_et_pb_old_content":"","_et_gb_content_width":"","_jetpack_memberships_contains_paid_content":false,"footnotes":""},"categories":[24],"tags":[],"class_list":["post-102220","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-lawsite"],"jetpack_featured_media_url":"https:\/\/i0.wp.com\/xira.com\/p\/wp-content\/uploads\/2025\/01\/BloombergLawAnswers-Featured-1024x576-fWWHcR.png?fit=1024%2C576&ssl=1","jetpack_sharing_enabled":true,"_links":{"self":[{"href":"https:\/\/xira.com\/p\/wp-json\/wp\/v2\/posts\/102220","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/xira.com\/p\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/xira.com\/p\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/xira.com\/p\/wp-json\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/xira.com\/p\/wp-json\/wp\/v2\/comments?post=102220"}],"version-history":[{"count":0,"href":"https:\/\/xira.com\/p\/wp-json\/wp\/v2\/posts\/102220\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/xira.com\/p\/wp-json\/wp\/v2\/media\/102221"}],"wp:attachment":[{"href":"https:\/\/xira.com\/p\/wp-json\/wp\/v2\/media?parent=102220"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/xira.com\/p\/wp-json\/wp\/v2\/categories?post=102220"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/xira.com\/p\/wp-json\/wp\/v2\/tags?post=102220"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}