LA Courts AI Pilot: Judges Draft Rulings with Learned Hand

Revolutionizing Justice: Los Angeles Courts Embrace AI for Judicial Efficiency

The legal landscape is undergoing a significant transformation, with artificial intelligence increasingly making its way into the hallowed halls of justice. In a pioneering move, the Los Angeles County civil court system has initiated a pilot program, equipping judges with an AI tool named Learned Hand. This sophisticated software is designed to tackle the formidable challenge of burgeoning caseloads by rapidly summarizing extensive legal filings and even drafting tentative rulings, aiming to enhance judicial efficiency while navigating concerns over accuracy and impartiality.

This initiative comes at a critical juncture, as court systems nationwide grapple with overwhelming workloads. The deployment of AI in such a sensitive domain underscores a growing recognition of technology's potential to alleviate administrative burdens within the judiciary. However, it also ignites a crucial debate about the ethical implications and the delicate balance between technological advancement and the preservation of public trust in the legal process.

"Learned Hand" AI: A New Judicial Assistant for Judges

The AI software, aptly named Learned Hand – a nod to the famed American jurist – represents a significant step towards modernizing court operations. Launched last month, the program grants select Los Angeles civil court judges access to a tool capable of distilling hundreds of pages of legal motions and, crucially, using samples of a jurist's writing style to help formulate conclusions and draft preliminary rulings.

According to court officials, the AI functions as an augmented intelligence system, not a replacement for human judgment. Rob Oftring Jr., the court's chief spokesman, emphasized, "Judicial officers have long been supported by research attorneys and law clerks who assist with summarization, legal research, analysis and drafting assistance. This assistance does not supplant the judicial officer's independent role in decision-making." The integration of Learned Hand is positioned as an evolution of this traditional support structure, providing judges with a high-tech "judicial sous chef" to manage the immense volume of documentation inherent in modern legal proceedings. The software's creator, Shlomo Klapper, a former attorney and federal law clerk, highlights that Learned Hand is already in use by court systems in 10 states, including the Michigan Supreme Court for reviewing appeal applications.

Addressing the "Paper Blizzard": Why AI Now?

The impetus behind adopting AI in the courts is stark: an increasingly strained judiciary drowning in a "paper blizzard." Klapper notes the rising number of self-represented litigants who, with public access to AI models like ChatGPT, are filing more cases in civil court, exacerbating an already heavy workload. This influx creates a "tsunami" of cases that existing human resources struggle to manage.

The problem of overwhelming caseloads is not unique to Los Angeles. Courts across the globe face similar pressures, often leading to delays in justice. The promise of AI lies in its ability to process vast quantities of information much faster than humans, thereby freeing up judges to focus on the nuanced legal analysis and ethical considerations that demand human intellect. By automating the preliminary tasks of summarizing documents and drafting initial decisions, Learned Hand aims to reclaim precious judicial time, enabling more timely and thorough deliberations on critical cases. This strategic use of AI could be a vital component in scaling AI for everyone within public services.

Concerns and Guardrails: Maintaining Legal Integrity

Despite the potential benefits, the introduction of AI into judicial processes has not been without controversy. Los Angeles County Dist. Atty. Nathan Hochman voiced concerns, acknowledging AI's utility in repetitive tasks like assessing summary judgment motions but expressing apprehension about its role in generating rulings. He warned that an AI-generated tentative ruling could unduly influence a judge's position before their independent legal analysis is complete. An L.A. County judge, speaking anonymously, echoed this, fearing that even if not adopted, an AI output could become an "unconscious reference point," subtly biasing subsequent decision-making.

The issue of AI "hallucinations" – where AI generates factually incorrect or fabricated information – remains a significant ethical challenge. Past incidents, such as attorneys submitting filings with AI-generated fake citations, underscore the need for robust safeguards. Klapper asserts that Learned Hand employs extensive guardrails, including a "Deep Verify" fact-checking process. This system interrogates every sentence of a generated order, ensuring facts align with case law citations, which are hyperlinked for easy verification. "We don't just tell the judges to trust us," Klapper stated, emphasizing transparency and human oversight. The court mandates judges to review and edit all AI outputs, signifying a commitment to maintaining judicial integrity. Discussions around such systems often involve best practices for input, similar to best-practices-for-prompt-engineering-with-the-openai-api.

Here's a snapshot of the Learned Hand pilot program:

Pilot Program Scope and Future Implications

The current pilot program for Learned Hand is structured to extend into early 2027, with a contract value exceeding $300,000. It is primarily focused on the civil court system, where it reviews and summarizes a diverse range of motions, from summary judgments to class-action settlements. While the current scope excludes criminal courts, the contract hints at limited future applications in that domain, specifically for postconviction relief. This cautious, phased approach allows for thorough evaluation and adaptation before broader implementation.

The success of this pilot in Los Angeles could set a precedent for other large court systems grappling with similar challenges. It highlights a pragmatic approach to leveraging AI not as a replacement, but as a force multiplier for human expertise, particularly in resource-constrained environments. The insights gained from how judges interact with Learned Hand, its impact on case processing times, and the effectiveness of its guardrails against errors will be crucial for the broader adoption of AI in legal tech.

The Broader Debate: AI in the Justice System

The Los Angeles pilot program is a microcosm of a larger, ongoing debate about AI's role across various facets of society, especially within critical sectors like the justice system. While some view AI as a "force for good" that can drastically reduce human work hours on tedious tasks, others remain wary of its potential to undermine foundational principles of fairness and impartiality. A Reuters survey noted that over 70% of respondents believed AI could positively impact the legal field by reducing manual effort.

Klapper addresses public anxiety head-on, likening Learned Hand not to "Skynet" (the malicious AI from Terminator) but to "Jarvis" (Iron Man's helpful AI assistant). This analogy underscores the intention behind the tool: to serve and augment human capabilities, not to dominate or replace them. The overarching goal is to empower judges to manage impossible caseloads more effectively, allowing them more time for the thoughtful decision-making that defines their role. As AI continues to evolve, the challenge will be to develop and deploy these tools in a manner that maximizes their benefits while rigorously upholding the ethical standards and public trust essential to the functioning of justice.