AI × Law

Legal AI startup Harvey has raised $200 million in a new funding round co-led by GIC and Sequoia Capital, pushing its valuation to $11 billion. The investment brings the company's total funding to over $1 billion as demand for legal AI tools accelerates.

A group of prominent U.S. writers, including Pulitzer Prize winner Michael Chabon, filed a proposed class action against OpenAI in San Francisco federal court. The lawsuit alleges OpenAI copied their copyrighted books and plays without permission to train ChatGPT to summarize their works and mimic their writing styles.

Indian courts are grappling with AI-generated fake citations and synthetic precedents entering judicial decisions, prompting severe warnings from the Supreme Court that unverified AI citations could constitute misconduct. In response, High Courts across India and the Supreme Court have introduced draft regulations and strict guidelines mandating independent verification and prohibiting AI from substituting judicial reasoning.

Plaintiffs in Mobley v. Workday have filed a motion for class certification in California federal court, alleging Workday's AI recruiting tools cause systemic disparate impact across race, gender, age, and disability. The filing relies on expert analyses showing rejection disparities exceeding 18 standard deviations and timing patterns indicating automated rejections.

Two recent federal court rulings demonstrate conflicting approaches to whether using third-party AI tools waives attorney-client and work-product protections. In SDNY, Judge Rakoff ruled that inputting information into Claude waived privilege, while an Eastern District of Michigan magistrate judge held that a pro se litigant's AI interactions remained protected work product.

China's Supreme People's Court released new judicial guidance establishing that AI-generated facial or voice replicas infringe personality rights if the real person is recognizable in the output. The guidance also makes utilizing a person's voice without consent as training material an infringement, while establishing notice-and-takedown liability for generative AI service providers.

Checkbox has unveiled First Pass, a context-aware AI agent designed to review incoming contracts, invoices, and legal requests before they reach in-house lawyers. The tool runs contracts against internal playbooks, checks LEDES invoices against billing guidelines, and automatically routes or resolves high-volume routine work.

Former Sidley Austin executive committee chair Mike Schmidtberger has joined Norm Law LLP as chairman, an AI-native law firm powered by software company Norm Ai, which raised $50M from Blackstone Innovations. Norm Law's model sidesteps non-lawyer ownership rules by functioning as a lawyer-owned entity licensing technology from Norm Ai.

Disney Chief Legal and Global Affairs Officer Horacio Gutierrez warned staff that his department will become a "much smaller organization" as the company automates workflows and adopts new technologies. The internal memo highlights plans to leverage modern tools, alternative legal service providers, and self-service models to reduce costs and increase efficiency.

The UK Ministry of Justice released a one-year progress update on its AI Action Plan, highlighting the deployment of AI tools across courts, probation services, and tribunals. Achievements include scaling its speech recognition tool, Justice Transcribe, to over 12,000 probation officers and First-tier Tribunal judges, alongside rolling out enterprise versions of ChatGPT and Microsoft Copilot.

Am Law 100 firm Bradley has selected Legora as its primary, firmwide AI platform. The rollout provides more than 1,200 attorneys and staff with agentic AI tools for document review, legal research, and drafting integrated into their daily workflows.

AI-driven patent law firm Fearn raised $5.5 million in funding from investors including Kindred Ventures, a16z Speedrun, Designer Fund, and Essence VC. The capital will support firm expansion and the development of FearnOS, a proprietary drafting and client management system designed to drastically reduce patent drafting times.

Tech-focused law firm Munck Wilson Mandala has implemented Harvey AI across its practice areas after switching from another vendor. The nearly 100-lawyer firm is deploying the platform to construct timelines in banking litigation, review discovery in trade secrets disputes, and execute M&A due diligence.

Teachers' Venture Growth has invested $50 million in legal AI platform Harvey as an extension to its recent financing, raising the round's total to $600 million at a $15.5 billion valuation. Harvey recently surpassed $400 million in annualized revenue and now serves over 3,000 customers globally across law firms and corporate legal departments.

Everlaw's 2026 report found that 49% of legal professionals actively use generative AI, reaching adoption levels in four years that took cloud ediscovery 12 years to achieve. With 17% saving 5-10 hours weekly, half of respondents say AI has altered or will soon alter conventional billing practices, driving interest in alternative fee arrangements.

A comparative review of Latin American high courts highlights emerging standards for AI in judicial proceedings, ranging from strict sanctions to permitted auxiliary use. Colombia's Supreme Court ruled that filing unverified AI-generated citations constitutes willful blindness and fined an attorney, while El Salvador permitted AI use without sanctioning a pro se citizen who submitted raw chatbot text.

A lawsuit alleging xAI's Grok was used to generate sexually explicit deepfake images will proceed with the plaintiffs using pseudonyms. The court's decision allows the victims to pursue their claims against the AI company while protecting their privacy.

The Thomson Reuters Institute and National Center for State Courts published an updated edition of their report on evidentiary issues raised by AI. The guide outlines strategies for courts to handle deepfakes and authenticating machine-generated evidence, highlighting rule updates in states like Utah, Connecticut, and Florida.

The Ninth Circuit has upheld the dismissal of Digital Millennium Copyright Act claims brought against GitHub and Microsoft over their AI tool Copilot. The ruling provides a key legal precedent limiting copyright liability for tech companies developing AI coding assistants.

Microsoft has urged a Manhattan federal court to grant it a favorable judgment, arguing that using books and news articles to train the large language model behind its Copilot AI tool constitutes fair use under copyright law.