TLDR: Generative AI (GenAI) continues to present a significant challenge in legal filings due to its propensity for ‘hallucinations’ – generating plausible but false information. While legal tech providers are implementing advanced techniques like Retrieval-Augmented Generation (RAG) to mitigate this, complete elimination remains unlikely. The solution emphasizes the critical role of human oversight and improved lawyering to ensure accuracy and navigate the associated legal and reputational risks, even as GenAI adoption rapidly increases across the legal sector.
Generative Artificial Intelligence (GenAI) tools, despite their transformative potential, are still grappling with a pervasive issue known as ‘hallucinations’ within legal filings. This phenomenon, defined by Thomson Reuters as ‘responses that sound plausible but are completely false,’ involves AI producing demonstrably false or entirely fabricated content. The problem is significant, with general-purpose large language models (LLMs) reportedly hallucinating on legal queries at rates between 58% and 82% of the time. This has led to real-world consequences, including lawyers being reprimanded for submitting court filings that cite non-existent case law generated by AI services. Even major tech companies like Apple have faced issues, having to suspend an AI feature that produced inaccurate news summaries.
Legal technology providers, including industry giants like Thomson Reuters (parent company of Westlaw) and LexisNexis, are actively working to address this challenge. They claim to mitigate hallucination risks through sophisticated techniques such as Retrieval-Augmented Generation (RAG), which integrates LLMs with auxiliary legal databases. However, these bold proclamations of ‘hallucination-free’ legal citations have yet to be accompanied by empirical evidence. Experts suggest that completely eliminating hallucinations is inherently difficult, as LLMs fundamentally operate by predicting the next word in a sequence based on statistical patterns learned from their training data, rather than understanding factual truth.
For news organizations and content creators, the uncritical use of GenAI outputs carries substantial legal risks, including potential breaches of copyright, data protection, defamation, privacy, and consumer law. The media and legal industries find themselves in a similar predicament, facing reputational damage and diminished trust if associated with AI-generated inaccuracies.
Despite these serious concerns, the adoption of GenAI within the legal profession is rapidly accelerating. The Thomson Reuters Institute’s 2025 Generative AI in Professional Services Report, based on a survey of over 1,700 professionals across the U.S., U.K., and Canada, reveals a strong belief in GenAI’s future role. A striking 95% of respondents anticipate GenAI becoming central to their organization’s workflow within the next five years. Current adoption rates are also on the rise, with 26% of legal professionals already utilizing GenAI, a significant increase from 14% in 2024. The legal industry is demonstrating the strongest GenAI adoption among all surveyed professions, with 28% for law firms and 23% for corporate legal departments.
The primary drivers for this growing acceptance are the tangible benefits GenAI offers. Legal professionals are increasingly optimistic about its ability to save time and boost productivity. GenAI tools designed for legal work can gather and analyze vast amounts of complex legal data, such as precedents and case law, in mere seconds, far surpassing manual human capabilities. This efficiency can also uncover resources that lawyers might not have otherwise discovered, leading to more informed decision-making. By automating repetitive tasks like document review, summarization, and presentation collation, GenAI allows legal professionals to dedicate more time to higher-value activities such as case strategy, client interaction, and complex legal reasoning.
Also Read:
- Legal Professionals Increasingly Embrace AI: Thomson Reuters Report Unveils Transformative Impact and Future Outlook
- Bloomberg Law Unveils Comprehensive Report on AI’s Transformative Influence on the Legal Sector
Raghu Ramanathan, President of Legal Professionals at Thomson Reuters, highlights the transformative power of the technology, noting its ‘rapid evolution.’ Esther Bowers, Chief Practice Innovation Officer at Honigman, emphasizes that GenAI is ‘not replacing what lawyers do best’ but rather ‘could free up attorneys to do more of those very things.’ This underscores the prevailing sentiment that while AI offers immense advantages, human oversight and skilled lawyering remain indispensable to ensure accuracy, ethical compliance, and ultimately, the integrity of legal practice in the age of artificial intelligence. Clients are also increasingly interested in, and even demanding, that their legal firms incorporate GenAI into their operations.
