Reliability & Liability of AI Attorneys

Reliability & Liability of AI Attorneys

By Omid Valinasab
Ford, Walker, Haggerty & Behar LLP

As expected, the exponential growth of artificial intelligence (AI) will change the future of humanity.[1] But what will be the effect of AI on the legal aspects of human life? This article provides a high-level overview of AI and its use within law, beginning with a general overview of AI and then turning to how AI is being used by lawyers in their practice, including litigation, transaction, and administrative law.

What is AI?

AI machines are software programs that can engage in human-like thought processes such as learning, reasoning, and self-correction, which enables the AI to simulate human intelligence.[2] Hence, the use of an AI machine will be limited to where the completion of a task requires human intelligence. According to ChatGPT[3], AI has a “lack of common sense, limited domain expertise, ethical challenges, data dependency, lack of creativity and intuition, and interpretability, and emotional intelligence.”[4]

Although AI machines have the ability to simulate human intelligence, “there is no such thing as ‘true AI,’ a program that codes itself to evolve and adapt.”[5] AI machines are just trying to mimic human brains. They have brain-like functions using artificial rather than biological neural networks.[6]  

Generative Artificial Intelligence

Generative AI describes algorithms that can be used to create new content, including audio, code, images, text, and videos. An example of generative AI is the well-known ChatGPT, a tool that answers almost any question. Although the answers are not necessarily the best answers at the moment, however, there is potential for more machine learning as time goes by for this AI. It is becoming more clear job performances are going to be affected by generative AI tools like DALL-E (AI-generated art tool) and ChatGPT. The risks and complications, however, are still unclear. Therefore, despite some negative reactions, it is fair to accept that machine learning clearly has the potential for good. Recently Thomson Reuters announced Generative AI is coming to “Westlaw Precision.” Thomson Reuters promised its Westlaw users to “soon be able to find the answers [they] need faster and easier than [they] ever thought possible.”[7]

Artificial Intelligence and Law

AI has emerged as a transformative force across various industries, and the field of law is no exception. The integration of AI technologies in the legal sector has the potential to revolutionize the way legal professionals carry out their work, streamline processes, and enhance efficiency. From legal research and contract analysis to predictive analytics and even automated document generation, AI applications are reshaping the landscape of legal practice. As noted by Mangan and Levin in their study on AI in the legal field, “AI technologies offer unprecedented opportunities to augment legal professionals’ capabilities and deliver more efficient and accurate legal services.”[8] This paper explores the implications, challenges, and future prospects of AI in the context of law.

  • The Good
    AIs can become beneficial to the lawyers as they can enhanced Legal Research, perform Contract Analysis and Due Diligence, and assist in Predictive Analytics. Firstly, AI-powered tools can analyze vast amounts of legal data, including cases, statutes, and regulations, to provide comprehensive and relevant insights. This improves the efficiency and accuracy of legal research. As noted by Mangan and Levin, AI systems “can conduct searches, extract and analyze relevant legal information, and generate summaries and recommendations, saving considerable time and effort for legal professionals.”[9]

    Secondly, AI algorithms can review and analyze contracts, identifying key provisions, risks, and anomalies more quickly and accurately than manual methods. This expedites due diligence processes and helps legal professionals make informed decisions. According to a study by Agrawal and colleagues, “AI tools can reduce the time and effort involved in contract analysis and due diligence tasks by automating the extraction and interpretation of relevant contract clauses.” [10]

    AI models can also analyze historical data and patterns to make predictions about case outcomes, judge rulings, and legal strategies. This assists lawyers in making informed decisions and developing effective legal strategies. As highlighted by Doshi and colleagues, “Predictive analytics tools can provide lawyers with insights into likely case outcomes, helping them assess the merits of a case and make informed decisions about litigation strategies.”[11]
    • The Bad
      There are many potential dangers in using AI in the legal field. For example, there is a possibility of bias and fairness AI systems are trained on historical data, which can reflect biases present in the legal system. If not carefully addressed, these biases can be perpetuated and lead to unfair outcomes. As highlighted by Crawford and Schultz, “AI systems can inadvertently perpetuate existing biases present in training data, leading to unfair or discriminatory outcomes in legal decision-making.” [12]

      Similarly, AI outcomes can suffer from a lack of contextual understanding. AI models may struggle with understanding complex legal concepts and contextual nuances that require human judgment. Legal reasoning often involves interpreting statutes, case law, and legal precedents, which may require a deeper understanding of the legal domain. According to Solum, “AI systems typically lack the capacity for understanding the underlying meaning and context necessary for nuanced legal reasoning.[13]

      Lastly, the use of AI in law raises ethical and privacy concerns. Ethical considerations such as the transparency of decision-making processes, data privacy, and security. AI algorithms may process sensitive and confidential information, requiring careful safeguards to protect client confidentiality and prevent unauthorized access. As highlighted by Calo and Rosenblat, “AI systems in law raise concerns about privacy, data security, and the potential for unauthorized access or misuse of sensitive legal information.” [14]

    Conclusion

    In conclusion, the integration of AI in the field of law holds immense potential for transforming legal practice. By leveraging AI technologies, legal professionals can streamline processes, enhance efficiency, and deliver more accurate and informed legal services. Generative AI can produce content and answer straightforward legal questions from statute or landmark cases. In contrast, a generative AI, at this time, cannot compare and contrast material facts of a case at bar with current court decisions.

    However, the adoption of AI in legal field also brings forth a range of ethical, regulatory, and societal challenges that must be addressed. As aptly stated by Solum in his work on AI and legal reasoning, “The successful integration of AI into the legal profession will require careful attention to the ethical, social, and legal issues that arise.”[15] It is crucial for stakeholders to engage in ongoing discussions, develop appropriate frameworks, and ensure responsible and transparent use of AI to maximize its benefits while upholding the principles of justice and fairness.


    [1] Joe McKendrick, How Fast Is Artificial Intelligence Growing? Look At The Key Bellwethers, FORBES (Dec. 19, 2018, 12:26 AM), www.forbes.com/sites/joemckendrick/2018/12/19/how-fast-isartificial-intell…-bellwethers/?sh=2e5f8977474a.

    [2] Omid Valinasab, Big Data Analytics to Automate Patent Disclosure of Artificial Intelligence’s Inventions, 27 U.S.F. Intell. Prop. & Tech. L.J. 133 (2023).

    [3] GPT stands for generative pretrained transformer. See Lock, Samantha, What is AI chatbot phenomenon ChatGPT and could it replace humans?, The Guardian (December 5, 2022). ChatGPT is an artificial intelligence chatbot developed by OpenAI and launched on November 30, 2022. It is notable for enabling users to refine and steer a conversation towards a desired length, format, style, level of detail, and language used. Successive prompts and replies are taken into account at each stage of the conversation as a context. While its content spans various domains of knowledge.

    [4] Retrieved Jun 23, 2023.

    [5] Valinasab, supra note 2.

    [6] Thaler, SL (1997). US Patent 5,659,666, “Device for the autonomous generation of useful information”, issued 8-19-1997. Washington, DC, US Patent and Trademark Office.

    [7] WestLaw, Thomson Reuters, Generative AI page  [last retrieved June 23, 2023], https://legal.thomsonreuters.com/en/c/westlaw/gen-ai-precision-waitlist.

    [8] Mangan, D., & Levin, D., Artificial Intelligence in the Law: The State of Play 2018. University of Cambridge Faculty of Law Research Paper No. 17/2018 (2018), https://ssrn.com/abstract=3148254.

    [9] Id at 124.

    [10] Agrawal, A., Greiner, M., & Quadrianto, N., What to expect of AI in contract analysis: The recent past and promising future. Artificial Intelligence and Law, (2019), 27(4), 527-537.

    [11] Doshi, A., Harsha, P., & Choudhury, G, Law in the Digital Age: The Role of Artificial Intelligence in Legal Decision-Making. In International Conference on Emerging Technologies in Data Mining and Information Security (2020), (pp. 282-289). Springer, Singapore.

    [12] Crawford, K., & Schultz, J. (2019). The limitations of fairness metrics in machine learning. In Proceedings of the Conference on Fairness, Accountability, and Transparency (pp. 477-486). ACM.

    [13] Solum, L. B., Artificial Intelligence and the End of Work. The Journal of Artificial Intelligence Research, 64, 231-240 at 237 (2019). https://www.jair.org/index.php/jair/article/view/11211.

    [14] Calo, R., & Rosenblat, A., The case for promoting AI literacy in law schools, Harvard Journal of Law & Technology, (2020) 33(1), 145-168.

    [15] Solum, supra, note 16.