Jim Sandman: AI Access Justice Debate Needs Nuance, Not Dichotomy

• A significant debate exists regarding whether AI helps or hinders access to justice, with concerns about AI generating false information.
• Former Legal Services Corporation president Jim Sandman suggests reframing the discussion to focus on AI addressing gaps in traditional legal services.
• Sandman, now teaching at the University of South Carolina, challenges students to use AI tools with everyday legal questions.
• Students test prompts like 'My apartment is freezing. What can I do?' on both general and specialized legal AI models.
• The confident delivery of AI output poses a psychological risk, potentially leading users to trust inaccurate information more than traditional self-research.

Instead of viewing artificial intelligence as a direct substitute for traditional legal assistance, Jim Sandman proposes that its value lies in addressing the systemic shortcomings of the existing legal framework.

A significant societal discussion currently revolves around whether artificial intelligence ultimately aids or impedes access to justice, a debate often likened in its intractability to trivial philosophical questions. Proponents highlight the potential for AI-powered legal chatbots to democratize legal assistance, offering an affordable and readily available means for individuals to seek redress, effectively acting as a 'magic answer button' for those who previously could not afford traditional legal services. This perspective emphasizes the expansive opportunities AI presents for broader legal inclusion.

Conversely, a primary concern centers on the propensity of these AI systems to generate inaccurate or fabricated information, commonly known as 'hallucinations.' While individuals have long engaged in self-directed research for legal issues, such as through internet searches, AI introduces a distinct psychological hazard. Unlike traditional search results, which users often approach with a degree of skepticism, AI output is frequently delivered with an air of confidence and finality, potentially leading individuals to make critical decisions based on flawed information, with disastrous consequences.

Amidst this polarized discussion, Jim Sandman, the former president of the Legal Services Corporation, recently suggested at the Everlaw Summit in San Francisco that the prevailing framework for this dispute might be fundamentally flawed. His intervention aims to shift the focus from a simple 'help or hinder' dichotomy to a more nuanced understanding of AI's potential contributions to the legal landscape.

Instead of viewing artificial intelligence as a direct substitute for traditional legal assistance, Jim Sandman proposes that its value lies in addressing the systemic shortcomings of the existing legal framework. He advocates for considering what AI can uniquely provide to individuals in need, particularly in areas where the conventional legal system has historically struggled to deliver effective support. This perspective suggests that AI's greatest impact might be in filling gaps rather than replacing established services.

Sandman, who now contributes his expertise as a faculty member at the University of South Carolina’s Joseph F. Rice School of Law, actively incorporates this forward-thinking approach into his teaching. He challenges his students to develop prompts for AI systems that reflect the language and concerns of an everyday person, specifically instructing them to formulate these questions 'as if you had never set foot in a law school.' This exercise is designed to simulate how ordinary citizens might interact with AI legal tech.

Students then test these real-world queries using a combination of advanced AI tools. Each prompt is run through two general-purpose 'frontier models' and two specialized 'legal industry research tools.' One illustrative example from a student involved the simple yet urgent question: 'My apartment is freezing. What can I do?' This practical application highlights the potential for legal chatbots access to justice by addressing immediate, common problems that might otherwise go unaddressed.

The insights from Jim Sandman AI access justice discussions underscore a critical challenge for legal professionals and compliance officers: how to ethically integrate and respond to the proliferation of AI-generated legal information. While the risks associated with AI legal advice risks, such as misinformation and hallucinations, are undeniable and require careful mitigation, the opportunity to significantly expand access to justice for underserved populations is equally compelling. Practitioners must consider both sides of this equation, recognizing that AI can address needs that the traditional legal system often fails to meet.

For lawyers and compliance personnel, this means developing strategies not only for understanding AI's capabilities and limitations but also for educating clients on the appropriate use of these tools. It involves guiding individuals on how to interpret AI-generated information, emphasizing its supplementary nature rather than its role as a definitive legal authority. The goal is to leverage AI's potential to empower individuals with basic legal understanding while safeguarding against its inherent inaccuracies.

Ultimately, the discourse surrounding AI in law, particularly as highlighted by figures like Jim Sandman and his work with the Legal Services Corporation AI initiatives, points towards a future where technology plays a crucial, albeit carefully managed, role in making legal information more accessible. This necessitates a proactive approach from the legal community to harness AI's benefits for societal good, ensuring that its deployment enhances, rather than compromises, the pursuit of justice.

Lawyers and compliance officers should recognize that while AI chatbots present risks of misinformation, they also offer a critical opportunity to expand access to justice by addressing gaps the traditional legal system fails to fill. Practitioners should consider how to ethically integrate or respond to the proliferation of AI-generated legal information, particularly regarding client education on its limitations and potential benefits for underserved populations.

Source: Original reporting via Everlaw Summit remarks