The lawyer's field guide to a fast-moving field.

California’s Senate Bill 574 establishes statutory rules governing how legal professionals use generative AI, prohibiting non-delegation of legal practice, mandating output verification, restricting input of confidential data, and requiring disclosure of AI use in court filings. The law also targets arbitrators and directs the Judicial Council to revise standards for judges, making California a pioneer in state-level legislative AI regulation for the legal industry.

Bloomberg Industry Group has acquired Regology, an AI-powered regulatory intelligence platform. The deal integrates Regology's regulatory monitoring and compliance workflow tools directly into Bloomberg Law's legal research and analysis services.

At its Inspire conference, NetDocuments positioned its cloud data store as the central governance hub for legal AI, using the Model Context Protocol (MCP) to integrate directly with third-party tools like Harvey and OpenAI models. The company aims to reduce consumption-based AI costs by grounding frontier models in law firms' internal context while maintaining strict data security controls.

Legal AI startup Harvey saw its gross margin plummet from 50% to -50% earlier this year due to flat-rate seat pricing combined with a 20-fold increase in token usage. The company restored positive margins in August by launching a proprietary model built on open-weight architecture, cutting cost-per-query on key features by 90%.

Employment law giant Littler has formed strategic partnerships with legal technology platform Legora and implementation provider Epiq. The collaboration aims to help corporate clients navigate the legal, governance, and operational impacts of AI-driven workplace transformations.

A report on in-house legal teams using Harvey found that 92% saw faster turnaround times and 88% spent more time on higher-value work. Case studies show energy company Repsol brought a mid-sized M&A transaction in-house to save an estimated €200,000, while Carrefour Spain deployed legal self-service tools for business units.

AI legal tech platforms Harvey and Legora have launched dedicated intake tools to capture, structure, and route incoming enterprise legal requests. The move marks a shift into legal front door technology, competing with specialized providers like Coheso.

Global law firm Goodwin has launched Goodwin AI Governance, a program pairing its attorneys with internal AI tech to help clients manage risk and regulatory shifts. The offering includes an annual subscription service called the AI Decision Desk, providing real-time monitoring of AI regulations and advice from Goodwin lawyers.

Brazilian investment firm Ventoá has purchased a 30% stake in Factum IA. The legaltech platform uses artificial intelligence to help trace debtor assets in judicial recovery and execution proceedings.

Law firm innovation leaders report that Anthropic is taking a rigid approach to contract negotiations, declining custom privacy and data security terms without high minimum spends. ILTA survey data shows that while 44% of firms use or explore Claude, only 7% have fully deployed it, as law firms cite a lack of legal-specific enterprise security and dedicated training support.

Ethem Coban, a former researcher at Meta's Superintelligence Labs focused on legal reasoning models, has left the tech giant to launch a legal AI startup. Coban is reportedly in early talks to raise $6 million at a $25 million valuation to help law firms and corporate legal teams build customized in-house models.

Trust and estate software startup EstateMin raised €1 million (approx. $1.1 million) in pre-seed funding. The company plans to use the capital to expand its operations across the U.S. and grow its go-to-market team.

Legal tech platform Legora (formerly Leya) raised $550 million in Series E funding led by Atreides Management, bringing its valuation to $8.8 billion. The company reported over $200 million in ARR across 140,000 legal professionals and highlighted its integration of Anthropic's Claude and Model Context Protocol connectors to protect client confidentiality.

The Chairman of Russia's Supreme Court, Igor Krasnov, met with Chinese officials in Beijing to request technology transfers and knowledge sharing on judicial AI integration. Russia plans to automate document checking, case assignment, and draft ruling generation across nearly all courts by 2030, leveraging China's existing 'Smart Court' systems.

A Fifth Circuit appeals panel has overturned a decision regarding a school diversity policy after identifying errors that appeared to stem from AI-generated hallucinations. The ruling highlights growing judicial skepticism and scrutiny surrounding AI-generated legal arguments and citations.

New class action suits against Alphabet and Duolingo highlight an escalating trend of 'AI-washing' litigation targeting public companies. Plaintiffs allege both companies misled investors regarding AI product release timelines and the quality impact of AI-generated content on core operational metrics.

The US Court of Appeals for the Fifth Circuit vacated an injunction and ordered a Mississippi case reassigned to a different judge after US District Judge Henry Wingate acknowledged his clerk used an AI tool, Perplexity, which resulted in errors in an order. The underlying litigation involves a Mississippi law banning public education funds for DEI programs.

A new survey by the Thomson Reuters Institute and the National Center for State Courts shows state courts shifting from evaluation to adoption to combat staffing shortages. Half of surveyed court professionals reported their court has established an AI policy or governance committee, while 25% plan to integrate AI into workflows within the next year.

Plaintiffs have launched five separate California federal lawsuits against Meta, alleging the company used pirated shadow libraries and a mass physical book-scanning operation to train its AI models. Counsel at Freedman Normand Friedland and Stris & Maher report representing over 2,200 authors with plans to file up to 1,000 additional individual cases.

U.S. District Judge Eumi Lee heard daylong summary judgment arguments in the copyright infringement lawsuit brought against Anthropic by major music publishers over training Claude on song lyrics. The judge took the case under submission without ruling, pressing both sides on market harm and whether the dispute could become the first AI copyright case to go to trial.