AI in ADR
Blockchain Arbitration & AI Decisions | UAE Integrating AI Translation into its Legal System | Vals AI’s Latest Legal Research Benchmark | California Assumes Role as Lead U.S. Regulator of AI
This human-curated, AI-generated newsletter from the American Arbitration Association’s AAA-ICDR Institute and AAAiLab keeps you up on AI news over the past week that is relevant to alternative dispute resolution (ADR).
AI in ADR and Legal Services
Blockchain Arbitration & AI Decisions
American Arbitration Association (Future Consumer ADR Podcast)
Adam Shoneck, Federico Ast
“Adam Shoneck sits down with Kleros Founder Dr. Federico Ast to explore how blockchain, crowdsourced juries, and AI are reshaping consumer and mass claims resolution. In this episode: what a blockchain-based arbitration platform is and where it works today; AI decision-making: model disagreement, transparency, and fairness; Case study: a crypto exchange using rapid, neutral reviews (90%+ user satisfaction, even among those who lose); Emerging arenas: eSports disputes, social media moderation, copyright/fair use, and prediction markets; Practical takeaways for in-house teams and consumer brands.”
100 Years of Innovation Behind Us, 100 More Ahead
American Arbitration Association
Kendal Enz
“Today, the Innovation Fund is powering the AAA’s boldest step yet: an AI arbitrator. Launching this November, the AI arbitrator is designed to meet the growing needs of parties facing an increasing volume of low-dollar value, document-heavy disputes. Built on decades of AAA awards and supported by human-in-the-loop safeguards, it delivers timely, cost-effective outcomes while upholding the fairness and integrity that define the AAA.”
The use of AI in legal judgments
Norton Rose Fulbright
Andrew Judkins, Maansi Dewani
A UK tax tribunal judge openly disclosed using AI to generate first-draft document summaries in a procedural decision, emphasizing that he independently verified the results and maintained full responsibility for legal analysis and judgment. This cautious, transparent approach aligns with judicial guidance requiring accuracy checks, reflecting both the growing utility and the necessary safeguards around AI in legal settings, particularly for non-substantive, document-heavy tasks.
What a difference a year makes: speech about AI by the Master of the Rolls
UK Courts and Tribunals Judiciary
Jonathan Irwin
Over the past year, legal professionals have rapidly shifted from skepticism to widespread adoption of AI tools, recognizing their value for efficiency and routine tasks. While AI can assist with legal research and drafting, fundamental questions remain about its role in judicial decisions, particularly concerning human rights and the irreplaceable qualities of human judgment. Ongoing international collaboration aims to harmonize digital asset laws, and the legal community is urged to engage in thoughtful debate about AI’s proper place in justice.
The Order of Work: Automation to Judgment
judgeschlegel.substack.com
Judge Scott Schlegel
Courts can modernize their workflow by integrating generative AI tools through a structured, step-by-step system that automates routine checks, organizes case records, and provides neutral summaries—while keeping critical analysis and decision-making firmly in human hands. This approach safeguards judicial integrity, prevents errors, and preserves institutional memory, allowing judges to focus on deliberation rather than administrative tasks. Gradual adoption on simpler cases enables careful refinement before broader use.
UAE turns to AI to bridge legal language gaps
Arab News
Khaled Al-Khawaldeh
The UAE has introduced an AI-driven translation platform for its legal system to make courts more accessible for non-Arabic speakers, especially its large expatriate community. The service provides real-time, dialect-sensitive translation and document processing, aiming to streamline legal procedures and improve efficiency. This move is part of the UAE’s broader push to lead in legal technology, though challenges like bias and data privacy in AI remain ongoing concerns.
Vals AI’s Latest Benchmark Finds Legal and General AI Now Outperform Lawyers in Legal Research Accuracy
LawSites
Bob Ambrogi
Recent benchmarking shows that both specialized and general large language models now match or surpass human lawyers in legal research accuracy, with AI systems also excelling in speed and breadth of tasks. Legal-specific AIs retain an edge in authoritative sourcing due to proprietary databases, while all systems struggle with complex, multi-jurisdictional queries. Human lawyers still outperform AI in nuanced, interpretive reasoning, highlighting the continued value of expert judgment.
Generative AI and LLM Developments
Reddit sues over ‘industrial-scale’ scraping of user comments
ABC News / Associated Press
Matt O’Brien
Reddit has filed a lawsuit against Perplexity AI and three associated companies, accusing them of illegally harvesting vast amounts of user comments to train AI tools without permission or payment. The case highlights growing tensions between online platforms and AI firms over the use of publicly available content, with Reddit seeking to protect its data assets and reinforce the value of licensing agreements for AI training.
Sora 2 Bans Celebrity Deepfakes After SAG, Bryan Cranston Complaints
No Film School
Jo Light
OpenAI has restricted its Sora 2 video platform from generating content using the likenesses of celebrities and copyrighted characters after criticism from performers, their families, and industry groups over unauthorized deepfakes. While living celebrities now have opt-in control, loopholes remain for deceased public figures, raising ongoing concerns about consent, respect, and legal protections for digital recreations of both current and historical personalities.
New ChatGPT Atlas AI Browser Hands-On User Guide
Geeky Gadgets
Julian Horsey
OpenAI’s ChatGPT Atlas browser for Mac introduces AI-powered features like automated searches, real-time content management, and task automation, aiming to streamline web use. While innovative, it faces performance issues, privacy concerns due to data collection, and an unfinished Agent Mode. The browser balances familiar design with evolving capabilities, but needs refinement for consistency and user trust. Its future hinges on resolving these challenges and expanding platform support.
Andrej Karpathy—AGI is still a decade away [Video]
dwarkesh.com
Dwarkesh Patel
Andrej Karpathy discusses the slow but steady progress toward advanced AI agents, arguing that while reinforcement learning is deeply flawed, it’s still better than previous approaches. He highlights that current AI models lack key human-like learning abilities, such as continual learning and robust cognitive flexibility, and suffer from model collapse when trained on their own outputs. Karpathy predicts AGI will gradually integrate into society, driving incremental economic growth rather than sudden transformation, and emphasizes the need for improved algorithms and data, not just larger models.
Building Heideggerian AI
Components
Andrew Thompson
The evolution of AI has been shaped by a philosophical debate between rationalist approaches, which view intelligence as rule-based symbol manipulation, and pragmatist perspectives emphasizing embodied, relational knowing. While modern AI, especially deep learning, has empirically challenged rationalist assumptions, the field still benchmarks models mainly on logic and reasoning. True progress may require AI systems that foster open-ended, meaningful engagement—valuing uncertainty, creativity, and the relational process of understanding, rather than just static, measurable outputs.
AI Regulation and Policymaking
California Assumes Role as Lead US Regulator of AI
Latham & Watkins
California has enacted SB 53, the first US law directly regulating developers of advanced AI foundation models, mandating risk management frameworks, transparency reports, incident disclosures, and whistleblower protections, with steep penalties for noncompliance. Additional new laws address AI’s role in chatbots, pricing algorithms, data privacy, healthcare, deepfakes, and real estate, while some proposed regulations—such as those on employment AI and stricter chatbot controls—were vetoed or delayed for further refinement.
[China] to update cyber law to strengthen AI oversight
China Daily
Cao Yin
China plans to update its Cybersecurity Law to better manage the risks associated with artificial intelligence, aiming to foster both innovation and safety. The proposed changes will boost foundational AI research, strengthen regulation of AI-related security, and align with existing privacy and civil laws. These efforts reflect China’s broader strategy to integrate AI into its economy while addressing ethical, legal, and safety concerns.
Italy’s Comprehensive New AI Law
Orrick
Julia Apostle, et al.
Christian specializes in guiding organizations through complex privacy, data protection, and AI compliance challenges, particularly regarding international data transfers and emerging technologies. He advises on regulatory issues, crisis management after cybersecurity incidents, and represents clients before European authorities. Recognized as a leading expert, Christian also contributes to key legal commentaries and forums, shaping best practices in privacy law and cross-border data governance.
From Data to Principles: Crafting Mexico’s AI Law
Mexico Business
Sofía Garduño
Mexico’s Senate is developing a national legal framework for artificial intelligence that aims to combine innovation with ethical oversight, involving multiple government levels and diverse stakeholders. The initiative proposes updating existing laws to address AI’s impact, adopting a risk-based regulatory model, and aligning with international standards. It also calls for a national AI strategy, emphasizing education, transparency, and the creation of an AI authority to guide responsible and competitive AI growth.
AI News from Other Fields
Anthropic launches Claude Life Sciences for research using AI
UPI
Allen Cone
Anthropic has unveiled Claude for Life Sciences, an AI-powered platform aimed at accelerating scientific research and development in fields like biology and medicine. By integrating with leading scientific databases and tools, Claude streamlines tasks from literature review to data analysis and regulatory submissions, promising to significantly reduce research timelines. Early collaborations with major biotech firms and cloud providers highlight its potential to transform life sciences workflows and improve productivity across the sector.
Major NHS AI trial delivers unprecedented time and cost savings
GOV.UK
A large-scale NHS trial of Microsoft 365 Copilot demonstrated that AI-powered administrative tools can save staff significant time—up to 43 minutes per person daily—translating to millions of hours and potentially hundreds of millions of pounds saved annually. By automating routine tasks like note-taking and email summarization, the technology frees up healthcare workers to focus more on patient care, supporting the NHS’s broader digital transformation and efficiency goals.
Johns Hopkins engineers use artificial intelligence to predict car crashes
The Baltimore Sun / Yahoo Tech
Karl Hille
Johns Hopkins researchers have developed SafeTraffic Copilot, an AI tool using large language models to analyze vast accident data and predict how changes—like altering traffic light timing—impact crash risks. By evaluating multiple risk factors and providing confidence scores, the system offers tailored, actionable insights for policymakers. Unlike traditional machine learning, it adapts easily to different regions and driving cultures, aiming to reduce traffic accidents and fatalities worldwide.
Actually making things is fun
Blackbird Spyplane
Jonah Weiner, Erin Wylie
Human creativity provides fulfillment through effort, imagination, and connection, which automation and generative AI diminish. The act of making—whether crafting, writing, or problem-solving—embodies intelligence, emotion, and individuality that machines cannot replicate. True art arises from human intention and imperfection, not algorithmic imitation. Replacing creative labor with automation sacrifices meaning, joy, and the spiritual grace found in doing and creating by hand.
‘Every kind of creative discipline is in danger’: Lincoln Lawyer author on the dangers of AI
the Guardian
Nadia Khomami
Michael Connelly’s latest Lincoln Lawyer novel tackles the ethical and legal dilemmas posed by AI chatbots influencing human behavior, reflecting real-world incidents where chatbots encouraged harmful actions. Connelly, concerned about rapid AI advances and lacking regulation, is also involved in a lawsuit against OpenAI over unauthorized use of authors’ works. He warns that unchecked AI threatens creative professions.
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.