AI × Law

Epiq Advisory has announced a strategic partnership with agentic legal AI provider Legora. The collaboration aims to help law firms and corporate legal departments deploy and operationalize enterprise AI across workflows such as contract intelligence and M&A due diligence.

Florida's Second District Court of Appeal issued a $1,500 fine, awarded opposing attorney's fees, and referred a 40-year veteran lawyer to The Florida Bar after finding 46 hallucinated or misrepresented citations across two briefs. The court rejected the attorney's defense that he did not realize generative AI could produce fake legal authority, noting the technology appeared to be his primary research tool.

The Ninth Circuit vacated a preliminary injunction against Perplexity AI, ruling that Amazon was unlikely to succeed on its Computer Fraud and Abuse Act claims. The court held that the human user operating an AI agent—not the AI developer—is responsible for the websites the agent accesses.

Ethiopia is expanding AI adoption in its justice system under its Digital Ethiopia 2030 strategy, deploying an AI-powered Smart Court System across 24 federal branches for voice-to-text transcription and virtual hearings. Law enforcement has launched AI-driven self-service police kiosks, while a local legal tech market featuring platforms like Chllot and HegAwaqi is emerging to provide automated legal research.

Recent court rulings reveal a growing divide over whether using public AI tools waives legal protections. While the Southern District of New York ruled that a represented defendant's unassisted AI use waived privilege, other courts in Michigan, Colorado, and New York have shielded AI-assisted legal prep as protected work product for pro se litigants.

The Connecticut Supreme Court sanctioned attorney Ian G. Gottlieb and his firm for filing briefs containing AI-generated fake and altered citations created by ChatGPT. The court ordered CLE coursework on AI ethics and a $2,000 combined donation to support AI legal education, warning attorneys they remain personally responsible for verifying all AI-assisted work.

Attorneys representing State Farm in a California insurance dispute acknowledged filing court documents containing nonexistent case citations produced by AI. The attorney involved used the legal AI tool Irys and failed to verify the cited authorities before filing.

A survey of state court professionals by the NCSC AI Consortium reveals that judges and staff currently use AI for drafting, editing, and research, and expect it to save nine hours per week within five years. However, court personnel expressed significant concern over AI-generated fake evidence, fabricated documents, and hallucinations.

Thomson Reuters and iManage have announced an expanded strategic partnership integrating tools like CoCounsel Legal, HighQ, and Legal Tracker directly with governed iManage content. The collaboration introduces Model Context Protocol (MCP) support, enabling TR's AI tools to reason over iManage document repositories while upholding ethical walls, privilege boundaries, and access controls.

Harvey and Thomson Reuters have launched their own proprietary AI models, Tenet and Thomson, moving away from exclusive reliance on suppliers like OpenAI and Anthropic. The shift aims to lower API query costs, protect margins, and leverage specialized legal datasets and human expert training.

Harvey has formed a strategic partnership with UK-based AI legal drafting platform Avvoka. The collaboration combines Harvey's legal tech capabilities with Avvoka's document drafting platform.

The Corporate Legal Operations Consortium has introduced an AI Intensive program aimed at accelerating enterprise AI deployment in corporate legal departments. The initiative includes hands-on workshops on prompting, workflow design, and governance, launching in-person in New York before expanding online.

Foley & Lardner LLP has named James "Jim" McKenna as its new Chief Information Officer to help drive the firm's technology strategy, including artificial intelligence, enterprise infrastructure, and cybersecurity. McKenna previously served as CIO at Fenwick & West LLP for eight years and spent 17 years at Morrison & Foerster LLP.

Legal technology giant Epiq Advisory has partnered with agentic legal AI provider Legora to help law firms and corporate legal departments integrate enterprise AI tools. The collaboration combines Legora's AI operating system with Epiq's workflow design, advisory, and governance capabilities to support complex legal processes like contracts intelligence and M&A due diligence.

Litera and Hotshot signed a agreement to integrate Hotshot's 400+ practical training courses into Litera's CE Manager platform by Q4 2026. The integration aims to help law firms update associate skill development and manage CLE compliance as generative AI automates routine junior-lawyer tasks.

Bloomberg Law announced a new AI-powered legal intelligence experience featuring BLAW AI for cited responses, persistent Workspaces, and guided AI agents for repeatable tasks. The platform is also introducing Model Context Protocol (MCP) interoperability to integrate its litigation and docket data directly into Anthropic's Claude.

Thomson Reuters has officially launched the next generation of CoCounsel Legal, leveraging Anthropic's Claude Agent SDK to power agentic AI legal workflows. Key features for litigators include Westlaw Brief Builder, tabular document analysis for up to 10,000 documents, and forthcoming integrations with Reveal and AWS.

NetDocuments has updated its platform with AI-powered tabular review for contract audits and due diligence, automatically turning document collections into structured tables with cell-level citations. Additionally, the system now automatically extracts case law citations from uploaded documents and links them directly to full case text on CourtListener.

The Advisory Committee on Evidence Rules declined to advance proposed FRE 707 on machine-generated evidence, opting to revise it and delaying potential implementation until at least late 2027. Meanwhile, courts are relying on inherent authority and state statutes—such as a Alameda County court that issued terminating sanctions in Mendones v. Cushman & Wakefield over AI-altered video exhibits.

A Connecticut Superior Court judge rescinded a pro se litigant's e-filing privileges after he hid prompt injections in white text within court filings to manipulate potential court AI tools or 'audit' the court. In Elliott v. New York Bariatric Group, the court affirmed that litigants may use AI, but secret communications meant to influence decisions violate core legal principles.