This human-curated, AI-generated newsletter from the 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
AAAi Podcast | Episode 11 | Hebbia.ai Head of Legal Vertical Ryan Samii
AAA-ICDR / Legally Disrupted
Bridget McCormack, Zach Abramowitz
“Bridget and Zach talk with Ryan Samii, Head of Legal Vertical at Hebbia.ai, about building the AI infrastructure legal teams actually want. He explains why Hebbia ditched chat for a grid—and how acting fast beats waiting for perfect. This episode is a playbook for smart risk, bold design, and legal teams who don’t wait to get started.”
Wolters Kluwer looks to the future of AI in arbitration with Open Innovation Hackathon
Wolters Kluwer
Wolters Kluwer and the AAA-ICDR are organizing an Open Innovation Hackathon on June 11 in The Hague. It will bring together legal and technology professionals to develop AI solutions for international arbitration. The event aims to accelerate innovation in dispute resolution by exploring practical AI applications that could streamline processes, reduce costs, and improve strategic outcomes, reflecting the growing influence of AI in transforming legal services.
120 court cases have been caught with AI hallucinations, according to new database
Mashable
Cecily Mauran
The legal field is experiencing a surge in AI-generated hallucinations, where chatbots like ChatGPT and Claude invent fake case citations, leading to errors in court filings. A growing database tracks dozens of such incidents, reflecting both the widespread adoption of generative AI for legal tasks and increased scrutiny from courts. While consequences for these mistakes have been relatively light, the trend raises concerns about reliability and due diligence in legal proceedings.
LexCheck 3.0 Delivers Unmatched AI Redlining Power
AP News / PRNewswire
LexCheck 3.0 introduces advanced AI-driven contract review that integrates seamlessly with Microsoft Word, delivering precise, context-aware redlines based on an organization’s negotiation history and preferences. The tool automates guideline creation, flags deviations, checks against historical precedents, and is engineered for both routine and complex, high-value contracts, offering scalable, enterprise-grade performance that streamlines legal workflows and enhances accuracy beyond previous solutions.
Generative AI and LLM Developments
Shots Fired: The AI Browser Wars Have Begun
Decrypt Emerge
Jose Antonio Lanz
A new wave of AI-powered web browsers—like Opera Neon, Dia, Surf, and Comet—are reimagining how users interact with the internet, offering features such as automated web app creation, intelligent tab management, context-driven organization, and advanced agentic search. Despite their technical sophistication and promise to address real user frustrations, these browsers remain niche products with minimal market share compared to dominant players like Chrome.
Anthropic CEO: AI will cut 50% of entry-level jobs for Gen Z
Windows Central
Kevin Okemwa
Leading tech figures warn that AI is set to disrupt white-collar employment, especially entry-level positions, potentially replacing up to half of such jobs. While some see opportunities for AI to address major health challenges, concerns persist about societal readiness for advanced AI and the risk of economic upheaval similar to past industrial shifts. Opinions differ on the immediacy and scale of AGI’s impact, but consensus points to significant workforce transformation.
The ‘white-collar bloodbath’ is all part of the AI hype machine
CNN
Allison Morrow
Anthropic CEO Dario Amodei recently warned that AI could soon eliminate half of entry-level office jobs and radically reshape the economy. Critics argue such claims serve more to hype AI’s power and inevitability, especially as Anthropic promotes its new chatbot. While generative AI can automate certain tasks, its limits and real-world impact remain unproven, raising skepticism about dramatic forecasts of economic upheaval.
FOD#102: Do Reasoning Models Think Too Much?
turingpost.substack.com
Ksenia Se
As AI models increasingly focus on complex reasoning, new research is emerging to make them more efficient by teaching them to avoid unnecessary or redundant thinking. Several recent approaches aim to help large language models adjust their reasoning depth dynamically, reducing computational costs without sacrificing accuracy. This trend signals a shift toward more self-aware and purposeful AI, raising questions about the future direction of model development beyond just efficiency.
What KPIs to Track When Deploying a Conversational AI Chatbot
Journal
Emily White
Effectively managing conversational AI chatbots requires ongoing attention to key performance indicators such as user engagement, conversation success, fallback and human handoff rates, customer satisfaction, resolution time, and goal completion. Monitoring these metrics helps organizations refine chatbot performance, ensure user satisfaction, and align automation with business objectives. Regular updates and expert guidance are crucial for maximizing both customer value and operational efficiency in a rapidly evolving digital landscape.
Chinese AI start-up DeepSeek pushes US rivals with R1 model upgrade
aol.com
Brenda Goh, Eduardo Baptista
Chinese AI firm DeepSeek has released an enhanced version of its R1 reasoning model, significantly improving its ability to handle complex tasks and reducing errors known as hallucinations. The update brings its performance closer to leading U.S. models while challenging assumptions about China’s reliance on expensive computing resources.
AI’s energy appetite is real—but so is its climate potential
TechRadar
Tim Weiss
While data centers powering AI consume rapidly increasing amounts of electricity, concerns about AI's environmental harm may be exaggerated. Innovations in energy-efficient hardware, greener power sources, and smarter algorithms are mitigating emissions. AI also enables significant sustainability gains by optimizing operations and supporting decarbonization efforts. Although AI's footprint is visible and growing, its potential to drive broader emissions reductions may outweigh its direct energy costs.
AI Regulation and Policymaking
Illinois House passes bill creating guidance for AI technology in schools, sends plan to Senate
WAND-TV
Mike Miletich
The Illinois House passed SB 1920, directing the State Board of Education to create guidelines on classroom AI—covering concepts, instructional uses, bias, privacy, transparency, and risk. The measure passed on a vote of 74-34 and returns to the Senate on concurrence.
Global AI Regulatory Update - May 2025
Eversheds Sutherland
Governments and regulatory bodies worldwide are rapidly advancing policies, guidelines, and infrastructure to manage the growing impact of AI. Central banks, financial institutions, and public agencies are adopting AI for efficiency and productivity, while new regulations and codes of practice emphasize responsible use, transparency, and risk mitigation. Initiatives span privacy protection, ethical standards, industry collaboration, and the creation of shared AI resources, reflecting a global push to balance innovation with oversight.
How to navigate global trends in Artificial Intelligence regulation
EY
Nicola Morini Bianzino, Marie-Laure Delarue, Shawn Maher, Ansgar Koene, Katie Kummer, Fatima Hassan-Szlamka
As artificial intelligence rapidly advances and becomes integral to sectors like healthcare, finance, and public services, governments worldwide are grappling with how best to regulate its use. Approaches vary, with some countries pushing for strict oversight while others favor minimal intervention. Balancing innovation with the need to address risks such as bias, misuse, and security challenges is now a central concern for policymakers across major global economies.
The Rogue AI Question: A Realistic Assessment of Risk in the Next Two Decades
vklewisauthor.substack.com
Vincent Lewis
As AI systems become more advanced, the main safety risks will likely stem from misalignment, unexpected emergent behaviors, and adversarial misuse rather than dramatic "rogue AI" scenarios. While progress in alignment and oversight is ongoing, challenges remain in detecting subtle failures and scaling safety measures. Proactive investment in research, regulation, and international collaboration is essential to minimizing harm and ensuring beneficial outcomes as AI becomes more integrated into critical domains.
AI News from Other Fields
Lessons (and an apology) from the Sun-Times CEO on that AI-generated book list
Chicago Sun-Times
Melissa Bell
After the Chicago Sun-Times published an AI-generated summer guide containing fabricated book recommendations, the incident exposed failures in editorial oversight and raised concerns about trust in local journalism. The event underscored the risks of using unvetted AI content, highlighted the need for clear AI policies and human accountability, and reaffirmed the importance of authentic, community-focused reporting amid financial pressures and technological change.
Politico’s Newsroom Is Starting a Legal Battle With Management Over AI
WIRED
Kate Knibbs
Politico’s newsroom union is challenging management over alleged breaches of their AI contract, claiming that new AI tools were introduced without required notice or negotiation and have produced content with ethical lapses and factual errors. This dispute could influence how much control journalists have over AI in newsrooms, as union contracts currently serve as one of the few enforceable mechanisms for AI accountability in journalism.
The Great Pivot: How Hollywood Studios Are Moving Beyond AI Experimentation
TheWrap
Jackie Snow
Hollywood’s attitude toward AI has shifted rapidly from skepticism to active adoption, with major studios like Amazon now openly integrating generative AI into production workflows. This change is driven by the availability of AI models trained on licensed data, addressing copyright fears, and by the need for efficiency as traditional production costs soar. The industry now sees AI not as a novelty, but as an essential tool for creative and economic survival.
An AI That Engages With Literature Without Destroying It: How Rebind.ai Surprised Me
PCMAG
Will Greenwald
Rebind.ai offers a unique approach to AI-powered literary analysis by combining large language models with curated commentary from renowned experts, providing users with direct, attributed insights into classic works. While the system occasionally produces awkward or contextually muddled responses due to AI limitations, its transparency and respect for original scholarship set it apart from typical generative AI tools, making it a promising, responsible enhancement for readers seeking deeper engagement with literature.
AI is helping rescue a nearly extinct bird species
Mashable
Amanda Yeo
A nonprofit arm of Colossal Biosciences has developed an AI-powered bioacoustics tool to identify calls of the critically endangered tooth-billed pigeon in Samoa, enabling conservationists to more precisely locate remaining birds and target protection efforts. Trained on scarce audio data, the algorithm distinguishes the rare bird’s calls with high accuracy and is being made open source, with hopes to extend its use to other threatened species and eventually empower citizen scientists.
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.