This human-curated, AI-generated newsletter from the American Arbitration Association’s AAA-ICDR Institute keeps you up on AI news over the past week that is relevant to dispute resolution and to legal services more broadly.
AI in ADR and Legal Services
AI Built For Law Outperforms ChatGPT, Claude, And Gemini On Legal Reasoning Benchmark
LawSites
Descrybe
Descrybe launched DescrybeLM, an AI system designed specifically for legal reasoning, which achieved a 100% accuracy rate on 200 multistate bar exam questions. In contrast, general-purpose AI models—ChatGPT 5.2, Claude Opus 4.5, and Gemini 3 Pro—missed between 13 and 23 questions, with DescrybeLM also scoring 99.70% in reasoning quality. The study highlighted that general-purpose models often provided confident yet incorrect answers, emphasizing the importance of purpose-built legal AI for reliable legal analysis.
The New Legal Frontier: AI Fuels Demand for Elite Tech Arbitrators
BriefGlance.com
George Millen
AI-driven disputes are rapidly increasing, prompting demand for specialized arbitrators with deep technical and legal expertise. SVAMC’s 2026 Tech List highlights elite neutrals addressing complex issues like IP conflicts, algorithmic bias, and AI liability. Specialized arbitration offers faster, confidential resolution, while new guidelines aim to ensure responsible AI use in legal proceedings.
The AI Productivity Paradox: Saving Time Takes Time
Mediate.com
John Lande
Professionals often feel too busy to learn AI tools that could enhance efficiency, leading to the ‘AI productivity paradox.’ This hesitation stems from the time required to learn and integrate AI, despite its potential to save time in the long run.
Smarsh Unleashes AI Agents to Reduce Corporate Legal Discovery Costs by 75%
Business Wire / Nexis
Smarsh has introduced AI-native agents that expedite legal discovery processes, reducing investigation timelines from weeks to hours and cutting costs by up to 75%. These agents identify relevant information early, reconstruct timelines, and visualize communication patterns across multiple languages. Additionally, Smarsh’s AI-driven surveillance capability enhances compliance oversight by detecting risk signals earlier and improving operational efficiency.
Generative AI and LLM Developments
What 81,000 people want from AI
Anthropic
In December, Anthropic conducted a global survey with over 80,000 participants from 159 countries to understand perspectives on AI. Respondents shared both positive experiences, such as AI aiding in accurate diagnoses, and concerns, like job displacement due to AI. The study highlights the complex relationship individuals have with AI, encompassing both hope and apprehension.
Sam Altman Admits That AI Is Disrupting the Basic Fabric of Capitalism
Futurism
Joe Wilkins
OpenAI CEO Sam Altman acknowledged that AI is fundamentally altering capitalism by shifting the balance between capital and labor. He noted that AI enables businesses to reduce labor costs, leading to a more abundant economy but diminishing worker power.
AI could trigger the biggest productivity boom ever
Big Think
Peter Leyden
AI is poised to revolutionize productivity by automating tasks, enhancing decision-making, and fostering innovation across industries. This technological advancement is expected to lead to significant economic growth and societal transformation.
Mistral AI launches Forge to help companies build proprietary AI models, challenging cloud giants
VentureBeat / Nexis
Mistral AI introduced Forge, an enterprise model-training platform enabling organizations to build, customize, and continuously improve AI models using proprietary data. This positions Mistral against major cloud providers in the enterprise AI market. Forge supports the full model training lifecycle, including pre-training on large internal datasets, supervised fine-tuning, and reinforcement learning pipelines.
Nvidia is launching NemoClaw, a more secure version of OpenClaw
The Verge
Stevie Bonifield
Nvidia introduced NemoClaw, an agentic AI platform that enhances the security and privacy of OpenClaw by operating within an isolated sandbox environment. Utilizing Nvidia Agent Toolkit software, NemoClaw optimizes OpenClaw with a single command and incorporates OpenShell to provide open models and enforce policy-based security, network, and privacy guardrails.
AI Regulation and Policymaking
Process over Product: How a California Bill Could Make Legal Services More Expensive
The Foundation for American Innovation
Blaine Dillingham
California Senate Bill 574 mandates that attorneys personally verify every citation in legal briefs, including those generated by AI, aiming to prevent AI-induced errors. This requirement could disrupt traditional law firm workflows, where tasks like citation-checking are typically delegated to junior attorneys or paralegals, potentially increasing costs for clients.
Analyzing the passage of state-level AI bills
Brookings
James S. Denford, et al.
In 2025, 260 AI-related measures were introduced in state legislatures, with 22 passing. Democratic lawmakers introduced about two-thirds of these bills, focusing on consumer protection, while Republicans often proposed bans on harmful AI uses and promoted innovation. States like California and Colorado pursued comprehensive AI governance, whereas others adopted targeted laws addressing specific issues like election deepfakes and AI-generated content warnings.
IRS Standards on AI and Tax Preparation Would Protect Businesses
Bloomberg Tax
Ryan Costello
The IRS is developing standards for the use of AI in tax preparation to safeguard businesses. These standards aim to ensure that AI tools used in tax preparation are accurate, reliable, and secure, thereby protecting businesses from potential errors and compliance issues.
By 2028, at least 80% of governments will deploy AI agents to automate routine decision-making, enhancing service delivery and operational efficiency. AI agents are autonomous or semi-autonomous software entities that use AI techniques to perceive, make decisions, take actions, and achieve goals in their digital or physical environments.
AI News from Other Fields
Steven Spielberg Explains Why He Hasn’t Used AI in Filmmaking
NewsBreak
Omar Gallaga
Steven Spielberg has not incorporated AI into his filmmaking process. He believes that AI lacks the emotional depth and understanding necessary to create compelling narratives, and that overreliance on technology could diminish the authenticity and emotional resonance of films.
AI Is Everywhere at SXSW, but Humanity Still Leads
ADWEEK
Kathryn Lundstrom
At SXSW, AI permeated discussions, with executives from 3M, Adobe, Accenture Song, Ancestry, and Monks sharing insights at ADWEEK House. They emphasized that integrating AI requires more than overlaying it on existing processes and highlighted the importance of ethical boundaries and human connection in marketing strategies.
The Robots Are Coming for Your Hair
Esquire
Andy Vasoyan
AI-powered apps are emerging to analyze and optimize hair care, offering personalized regimes to combat hair loss and promote overall hair health. These applications require users to photograph their scalps from specific angles, enabling the software to assess follicle count and condition.
US startup advertises ‘AI bully’ role to test patience of leading chatbots
The Guardian
Amelia Hill
Memvid, a California startup, is offering an $800-a-day position titled ‘AI bully’ to individuals who will spend eight hours interacting with AI chatbots, aiming to expose their inconsistencies and limitations. The role requires no technical expertise, only a history of frustration with technology and the patience to repeatedly challenge chatbots.
SAP to overhaul fee structure for the AI era
Reuters / MarketScreener
SAP is revising its fee structure to align with the AI era, focusing on enhancing AI capabilities and integrating AI into its offerings. The company is investing in AI-driven solutions and restructuring its operations to better serve customers.
The AAA-ICDR and AAAiLab do not endorse, control, or assume responsibility for the accuracy, availability, or content of any summaries, articles, or external sites or links referenced in this newsletter. Neither the AAA-ICDR nor AAAiLab are compensated or otherwise incentivized by any third parties for including content in this newsletter. All copyrights, trademarks, and other content referenced in the articles or article summaries remain the property of their respective owners. If you believe we have linked to material that infringes your copyright rights, please contact us at institute@adr.org, and we will review and address the matter promptly. This newsletter does not constitute legal advice.
The AAA-ICDR respects your privacy. If you no longer wish to receive these emails, you may unsubscribe at any time by clicking the link below.
We also do not share your personal information without your consent. For more information, please refer to the AAA-ICDR’s full privacy policy.