AI × Law

In the authors’ lawsuit over Meta’s alleged use of pirated books to train Llama, Judge Vince Chhabria has already granted Meta summary judgment on fair use, but only on the record before him. The docket entry here shows the case is still moving, with new discovery and case-management activity; the ruling’s caveat makes it important beyond Meta because plaintiffs in other training-data cases are using it as a roadmap.

Anthropic has filed a counterclaim for non-infringement in a copyright suit brought by book author Daniel Benjamin Gilbert. The move adds a new defensive tactic in AI copyright litigation and could signal how defendants try to reshape these cases.

OpenAI has quietly updated ChatGPT to refuse direct requests to mimic living or dead authors by name, including Hemingway, Stephen King, and Agatha Christie. The change appears designed to reduce copyright risk as author lawsuits and high-profile AI copyright rulings intensify.

NetChoice submitted comments urging the FTC to keep its AI policy statement focused on deceptive practices, draw a sharper line between model developers and deployers, and create safe harbors for good-faith safety guardrails and state-law compliance. The filing is a concrete signal in the fight over whether federal law will override a growing patchwork of state AI rules.

OpenText says its newly launched Aviator Flex feature lets legal teams run custom AI prompts across document sets and save the results into searchable fields. The company is pitching it for e-discovery, investigations, case strategy, and contract analysis, which could make AI outputs more reusable in litigation workflows.

British Labour MP Jess Asato filed a claim against xAI over Grok-generated non-consensual sexualized images of her, seeking a court order, damages, and deletion of the images and related data. The case alleges breaches of UK data protection law and misuse of private information, and it adds to growing legal pressure on AI image tools that generate deepfakes.

A second lawsuit was filed in Arkansas against xAI, accusing its programs of creating child pornography. The case adds to legal pressure on AI companies over harmful outputs and could test liability theories for generative AI systems.

In Conservation Law Foundation v. Shell Oil, a magistrate judge said an expert’s use of generative AI to narrow a document production was part of the expert’s methodology and could be subject to discovery under Rule 26(b). The court ordered revised responses identifying any prompts or queries used, and warned of sanctions if the representation was later shown to be false, though the order is stayed pending objection.

New Zealand Police are using AI and related analytics for crime detection, including BriefCam video analysis, Microsoft 365 Copilot Chat, and automated report triage. The article also flags ongoing legal challenges over facial recognition and ANPR, plus a New Zealand Supreme Court ruling that police photos uploaded without a specific investigative purpose were an unlawful search.

Patent Bots says it is rolling out and demoing a set of AI patent tools, including Gen AI Chat for drafting claims and specs, Gen AI Saved Prompts, rules-based generation, PatentPlex for prosecution and litigation research, and Premium Examiner Stats. The article also says the company has been around since 2018 and offers the platform via Word and browser with a 14-day trial.

Quinn Emanuel says it has built a practice focused on disputes tied to AI data center financing, construction, and operation. The firm points to litigation and arbitration over private credit structures, securitizations, GPU-collateralized debt, zoning, securities disclosures, and construction delays.

A consolidated class action in the Northern District of California accuses Otter.ai’s meeting assistant of recording calls, generating transcripts and voiceprints, and using that data to train its models without proper consent. The article says Otter’s motion to dismiss is fully briefed and oral argument is set for August 2026, making it a live test of how privacy, wiretap, and biometric laws apply to AI products that learn from user data.

A federal judge said the Trump administration has not justified labeling Anthropic a national security risk and suggested the company is likely to prevail in part of its case. The dispute matters because it involves government limits on AI vendor use and a direct clash over an AI company’s speech and due process claims.

The University of Tennessee Research Foundation filed what the article says is the first university lawsuit against an AI company, alleging Anthropic infringed patents covering artificial neural networks. The case could test whether universities can claim a stake in the core architecture behind large language models, not just the training data used to build them.

A federal judge approved Anthropic’s $1.5 billion settlement with authors, described as the largest copyright settlement in U.S. history and the largest copyright class action ever certified. The deal centers on allegations that books from shadow libraries were used to train Claude, and it underscores how AI firms can face massive exposure over data sourcing even as training itself may still be defended as fair use.

A long analysis of Alabama State Bar Formal Opinion 2026-01 says the bar treats AI as creating no new ethics duties, but as intensifying existing obligations around competence, confidentiality, supervision, candor, and verification. It also points to AI-sanctions cases, including the April 2026 Ibach v. Stewart appeal dismissal and fee award, as practical warnings for lawyers and legal departments.

The California State Bar has issued new 2026 guidance replacing its 2023 AI advice, covering generative AI and agentic AI. It says lawyers must keep control over judgment, verify citations, protect confidentiality, and disclose AI-related costs and limits where appropriate — a notable ethics roadmap as AI-related sanctions and discipline rise.

Mintz promoted John Koss to the firm’s first Chief Innovation & AI Officer. He will oversee firmwide AI strategy, AI investments, agentic technology initiatives, and internal and client-facing development, making this a concrete Big Law leadership move around AI adoption.

Legal AI company Legora has acquired London-based Wexler, whose software extracts, verifies, and structures facts from large document sets for litigation and investigations. The deal gives Legora a source-traceable fact layer for its agentic workflows and expands its AI capabilities for law firms and in-house teams, following Legora’s $600 million Series D at a $5.6 billion valuation.

Inspira, a Brazilian legal technology startup, raised a $2.8 million Series A led by Cloud9 Capital with participation from Vivo Ventures. The company says its AI assistant supports legal research, document analysis, drafting, and case management for lawyers, law firms, corporate legal teams, and government institutions.