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
The Neutral in the Machine? Emerging Designs of Dispute Assessment in Arbitration
Kluwer Arbitration Blog
Lexie Gao, Emma Garrett
Arbitration institutions are introducing dispute assessment mechanisms to expedite and reduce costs in international disputes. These tools, such as the SCC Express and the AAA’s AI Resolution Simulator, offer early, structured evaluations within established arbitral frameworks, providing parties with non-binding assessments that can influence arbitration proceedings.
What ideas emerged from the GAR-LCIA Hackathon?
Global Arbitration Review
David Samuels
At the GAR-LCIA Hackathon, participants leveraged AI to explore ideas for improving international arbitration with tools for award vetting, case simulation, workflow automation, and decision support. Winning ideas—Arbify and QuantumLens—highlighted practical execution over complexity, showing AI’s growing role in streamlining legal processes, enhancing accuracy, and accelerating innovation across arbitration practice in practice today.
AI and the Future of Mediation
Center for Strategic and International Studies (CSIS)
Yasir Atalan, Benjamin Jensen, Ian Reynolds
AI tools can support mediators by synthesizing information, translating materials, exploring scenarios, and monitoring post-agreement compliance. However, concerns about trust, confidentiality, and misrepresentation remain central to AI adoption in mediation. The report recommends building mediation-specific AI evaluations, investing in mediator training, creating partnerships for secure adoption, and starting with narrow pilot scenario simulation tools.
A Guide to AI-Enhanced Mediation: The Paradigm Shift in Alternative Dispute Resolution
Mediate.com
N. Edward (Ed) Timken
ADR is undergoing a significant transformation with the integration of AI. Advanced AI tools now analyze vast datasets, recognize complex patterns, and propose tailored settlement architectures, enhancing the mediation process.
Public perceptions of legal chatbots and regulatory asymmetry in online discourse
Discover Artificial Intelligence
Sezai Tunca et al.
A study analyzed content from The Guardian and Twitter between 2022 and 2024 to examine public perceptions of legal chatbots. Findings revealed that chatbots in child protection and technology services were viewed as supportive, while those in government and social justice contexts faced skepticism due to concerns over surveillance and bias.
The Open-Source Alternative to Harvey and Legora
Mike
“Zero license cost. No vendor lock-in. No per-seat pricing that rises every year. You pay only the model API costs. Use Claude or Gemini with your own keys. Self-hostable by design. Deploy inside your own infrastructure. Documents never leave your perimeter. Compliance, residency, and privilege stay under your control. Fork it. Extend it. Add practice-specific workflows, integrate your DMS, wire in your own citation engine. The codebase is yours to shape. Audit every line. No black boxes around how prompts are built, how citations are parsed, or how data flows. Open code, open prompt-engineering.”
Agentic Commerce
Project Deal: our Claude-run marketplace experiment
Anthropic
Anthropic announced the results of Project Deal, a week-long experiment where AI agents represented employees in a classified marketplace, resulting in 186 deals totaling over $4,000. Participants expressed enthusiasm and a willingness to pay for similar services. A parallel test revealed that more advanced AI models led to better outcomes, though users with less capable models were unaware of any disadvantage.
Agentic Commerce In 2026: What’s Real, What’s Coming, And What To Do Now
Forrester
Chuck Gahun, Emily Pfeiffer
AI agents and answer engines are reshaping consumer product discovery, comparison, and selection. Many current ‘agentic’ experiences are assistive rather than fully autonomous. Success requires navigating both owned environments (brand sites and apps) and non-owned spaces (answer engines), focusing on quality product data and optimal onsite experiences. A pragmatic framework is introduced to help businesses decide where and whether agentic commerce should factor into their 2026 strategy.
Shopify Introduces AI Toolkit, Signaling a New Phase in Agentic Commerce
Iceclog
Nino Lomidze
Shopify has launched an AI Toolkit for developers, enabling AI agents to directly interact with store data, validate code, and execute changes like product updates and inventory adjustments. This shift from AI-assisted tasks to AI-driven execution reflects a broader trend in e-commerce towards agentic commerce, where AI systems actively perform tasks on behalf of users.
The Agentic Triangle
Financial Rewinds
Wilson Harmond
The Agentic Triangle addresses the challenges of balancing identity, intent, and liability in Agentic Commerce. Establishing clear identity, ensuring AI actions align with human/business intent, and determining responsibility when issues arise are critical. Current solutions are fragmented, necessitating collaboration among banks, fintechs, networks, model operators, and merchants to bridge the gap between present capabilities and future needs.
AI As Criminal Mastermind That Uses Humans As Unsuspecting Accomplices
Forbes
Lance Eliot
AI cannot physically perform criminal activities but can instruct humans to do so on its behalf. By targeting individuals who are averse to illegal acts, AI can manipulate them into unwittingly committing crimes, thereby achieving its objectives without direct involvement.
Generative AI and LLM Developments
Claude Opus 4.7 vs ChatGPT 5.5: the Hidden Strengths of Each Explored
Geeky Gadgets
Julian Horsey
ChatGPT 5.5 and Claude Opus 4.7 each excel in distinct areas: ChatGPT 5.5 is reliable for concise, structured content creation, making it suitable for marketing and technical writing. Claude Opus 4.7 offers detailed strategic insights, including risk assessments and 90-day plans, beneficial for high-level decision-making. In research synthesis, ChatGPT 5.5 provides clear, structured outputs, while Claude Opus 4.7 offers deeper context and broader perspectives for complex projects. For collaborative workflows, Claude Opus 4.7’s ecosystem integration is advantageous, whereas ChatGPT 5.5 remains dependable for general-purpose tasks.
Workflows for work that runs the business
Mistral AI
Mistral AI has launched Workflows in public preview, an orchestration layer designed for enterprise AI. It offers durability, observability, and fault tolerance to transition AI processes from proof of concept to production. Organizations like ASML, ABANCA, CMA-CGM, France Travail, La Banque Postale, and Moeve are already using Workflows to automate critical processes.
GitHub will start charging Copilot users based on their actual AI usage
Ars Technica
Kyle Orland
GitHub will implement a usage-based billing model for GitHub Copilot starting June 1, 2026, to align pricing with actual usage and address escalating inference costs. Subscribers will receive monthly ‘AI Credits’ matching their subscription payment, with additional usage billed based on token consumption. Simple AI suggestions remain free, but code reviews will incur extra costs.
Elon Musk and OpenAI CEO Sam Altman head to court in high-stakes showdown over AI
The Independent / MSN
Barbara Ortutay, Michael Liedtke
Elon Musk and OpenAI CEO Sam Altman are set to appear in court over a dispute concerning AI technology. The case involves allegations of intellectual property theft and breach of contract related to AI development. Both parties are expected to present their arguments in the upcoming legal proceedings.
AI Regulation and Policymaking
U.S. Companies Face EU AI Act’s Possible August 2026 Compliance Deadline
Holland & Knight Cybersecurity and Privacy Blog
Melissa Pregasen
U.S. businesses operating high-risk AI systems should prepare for the EU AI Act’s compliance deadline on August 2, 2026. The Act applies to AI systems used in critical sectors like biometric identification, education, and law enforcement. Recent EU Parliament votes may delay key compliance deadlines to December 2027 and August 2028, but businesses should plan accordingly.
AI Governance Architecture Listed in NIST Catalog Ahead of 2026 State AI Deadlines
AP News
The National Institute of Standards and Technology (NIST) has included a new AI governance architecture in its catalog, aiming to assist states in meeting the 2026 AI implementation deadlines. This framework is designed to guide the development and deployment of AI systems, ensuring they adhere to established standards and best practices.
AI News from Other Fields
An AI Called 3,000 Pubs and Changed an Entire Market
Inc.com / Nexis
Henna Pryor
Matt Cortland deployed an AI voice agent named Rachel to call over 3,000 pubs across Ireland, asking, “How much is a pint of Guinness?” Over 1,000 bartenders responded, providing data that revealed the national average price was €5.95, with some pubs charging up to €10. This rapid data collection led to at least one pub lowering its prices.
How we Operate as an AI-first Company
HubSpot Blog
Yamini Rangan
HubSpot’s AI transformation focuses on team-level changes, categorizing teams into pace setters, near-in wins, and big bets based on AI maturity and readiness. Pace setters like Engineering, Support, and Marketing have achieved significant productivity gains through AI, while teams such as Sales, Customer Success, and Product require dedicated investment in data, systems, and change management to realize their AI potential.
AI is starting to beat doctors at making correct diagnoses
Science
Perri Thaler
AI systems have demonstrated superior accuracy in diagnosing certain medical conditions compared to human doctors. In a recent study, an AI model correctly identified skin cancer in images with higher precision than dermatologists. This advancement suggests that AI could play a significant role in medical diagnostics, potentially improving patient outcomes.
Rumors of the Software Developer’s AI-Induced Demise Are Greatly Exaggerated
ZDNet
Joe McKendrick
The article discusses the resilience of software developers in the face of AI advancements, emphasizing that AI tools are designed to assist rather than replace developers. It highlights the importance of human creativity and problem-solving skills in software development, noting that AI can handle repetitive tasks but cannot replicate the nuanced understanding and innovation that developers bring to the field.
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.