Q2 AI Disputes Monitor: 42 Lawsuits Filed, 35% Increase Over Q1

AI disputes have nearly doubled since last year. Ocean Tomo continues to deliver fact-based, well-researched insights on this trend. For more information, check it out: https://lnkd.in/gP6BgNkq #AI-disputes; #AI-lawsuits

AI adoption is moving faster than legal certainty. But what happens when the data powering your models—or the content they generate—carries hidden copyright risks? For most organisations, AI data provenance is a black box. Unlicensed training data, opaque third-party tools, and uncertain output rights are creating significant legal and financial exposure. With the EU AI Act now in motion, waiting is no longer a viable strategy. Enter ClearSource by Agentize. ClearSource is our comprehensive AI Data Provenance & Copyright Exposure Assessment. We map your entire AI supply chain—from internal models to third-party tools—to give your boardroom one decision-ready picture of your legal risk. More importantly, we don't just report the risk; we provide the prioritised, actionable steps to remediate it. Know your AI's legal exposure—before it becomes a liability. Protect your innovation. Book your ClearSource assessment today at agentize.eu. #Agentize #ClearSource #AI #ArtificialIntelligence #EUAIAct #DataProvenance #CopyrightLaw #TechLaw #AIGovernance #RiskManagement #Innovation

Build for the courtroom standard, not the demo standard. I opened this month writing about lawyers sanctioned over AI output they couldn't verify. I'm closing it with the principle that should prevent every one of those headlines. Most tools marketed as "AI for law firms" are search-and-summarisation engines wearing a compliance costume. Demo rooms love them. Clean interface. Impressive output. Neat highlights. Then comes the question that ends the demo: which specific rule did this document violate? Silence. Every time. Real compliance automation has one standard. This document fails. Here is the rule it violates. Here is the regulation citation. Here is precisely how many days late it is. No confidence score. No "based on documents similar to this one." The rule applies or it doesn't. The line I hold every time we build Docly: if it can't survive cross-examination in court, it's not ready for the legal industry. Agree or disagree? #legaltech #aicompliance #nofaultlaw #legalinnovation #innovation #documentvalidation

Holiday reading! Now on SSRN, a chapter written by Zoi Krokida, Ioanna Lapatoura and me on UK and EU liability in relation to generative AI under ***trade mark law*** in the forthcoming Cambridge Handbook of Artificial Intelligence & Trademarks (Bonardio, Lucci and Alonso eds) Expect -the lowdown on primary infringement for gen-AI providers post-Getty -secondary infringement -platform liability under the EU regimes https://lnkd.in/eJp4QnVn

You cite three cases. They look right. The citations are formatted correctly. The case names sound real. They are not real. This has happened. It is on the record. Lawyers have filed AI-generated briefs containing fabricated authorities, and courts have noticed. Some of those lawyers are still explaining themselves to their law societies. The cost is not just embarrassment. It is your reputation, your file, and depending on the day, your licence. Here is the quiet part most people skip: the problem is not that AI drafts. The problem is that AI drafts confidently. It does not flag uncertainty. It does not know what it does not know. And when you are billing six files at once, a citation that looks right usually gets through. The 2-minute check is not a workflow luxury. It is the minimum standard of care in a world where the tool you are using can hallucinate with complete conviction. Every citation Juris surfaces links to a real, verifiable authority. If it cannot confirm the source, it will not give you the citation. That is not a feature. It is the standard. What does your current verification step actually look like? #Juris #LegalTech @KwataTeam https://lnkd.in/gFCh4aSf

Today we start building your AI policy Before I get into the framework, one thing I want to be clear about: an AI policy is not a legal document. It doesn't need to be reviewed by a lawyer before it exists. It needs to be reviewed by a lawyer before it becomes binding on employees or clients in high-stakes ways but the version you write this week is a governance starting point, not a legal instrument. Treat it that way and you'll actually write it. Try to make it bulletproof before publishing it and you'll spend six months on a document that never gets used. If your firm handles regulated data — HIPAA, SEC, bar-covered, get your relevant counsel to review the final version before it goes live. For most professional services firms, that's one conversation, not a lengthy drafting process.

https://lnkd.in/e-926yK8 Explore how UK lawyers can leverage AI for summarizing contracts, pleadings, and case law. Discover tools, risks, and best practices for accurate and secure summaries. 🤖📄 #GoClio

New York just officially permitted AI in court filings. Effective June 1, 2026, Part 161 — the New York State Unified Court System's new AI rule, no longer prohibits attorneys from using AI to draft legal submissions. No mandatory disclosure. No checkbox. No separate form. Just one requirement: if you use AI, you personally guarantee every citation, every case reference, every statute is real. Your signature on the filing is your certification. Not the AI's. Yours. Courts across the country sanctioned attorneys over $145,000 in Q1 2026 alone for AI-generated fabrications. Nebraska suspended a lawyer indefinitely in April after 57 of his 63 citations were flagged as defective 20 were outright hallucinations. A Mississippi judge just canceled an entire trial after discovering both sides had used AI. New York's rule didn't create new liability. It codified the liability that already existed and made clear it travels with the attorney's signature, not the tool. This is the problem DoubleCheck Legal was built to solve. We connect businesses with certified attorneys who review AI-generated legal documents, catching hallucinations, fabricated citations, and legal errors before they become your attorney's liability problem. You get the speed of AI. The reliability of a real lawyer. → Learn more at https://lnkd.in/e3xq2Qrh #LegalTech #AI #AIinLaw #LawFirms #LegalRisk #DoubleCheckLegal #NewYork #Part161

Could Your AI Chats Become Evidence One Day? As lawyers and businesses increasingly turn to AI to summarise contracts, identify legal risks, review due diligence documents and even brainstorm litigation strategy, one question deserves far more attention than it currently receives: What happens to everything we share with AI? Many users treat AI as a private workspace. Entire agreements are uploaded for review, confidential board papers are summarised, commercial negotiations are analysed, and legal strategies are discussed. While AI undoubtedly improves efficiency, these interactions also involve sharing information with a third-party platform. Depending on the AI service being used, its contractual terms, data handling practices and the applicable law, this may raise important questions around confidentiality, privilege and data governance. Recent developments in the United States show that these concerns are no longer theoretical. Courts have begun examining whether conversations with AI platforms can be sought during litigation. In United States v. Heppner, a federal court held that documents created using a public AI chatbot were not protected by attorney-client privilege, reasoning that the communications were made to a third-party platform rather than to a lawyer. Other courts, however, have reached different conclusions in specific factual circumstances, particularly where AI functioned as a drafting tool for self-represented litigants. The law is still evolving, but one thing is clear: AI conversations are increasingly becoming part of legal disputes rather than remaining outside them. This has implications well beyond litigation. Businesses routinely operate under confidentiality obligations, non-disclosure agreements and trade secret protections. If employees upload sensitive contracts, customer information or internal strategy documents to public AI platforms without appropriate safeguards, organisations may expose themselves to contractual, regulatory and evidentiary risks. The issue is not whether AI should be used, but how it should be used responsibly. The lesson is not to stop using AI. It is to treat AI as any other third-party technology provider. Understand the platform's terms, implement internal AI governance policies, avoid uploading confidential information into public tools unless authorised, and distinguish between consumer AI platforms and enterprise solutions designed with stronger privacy and security controls. As AI becomes embedded in everyday legal and commercial workflows, organisations may need to start asking a new question before clicking "Send": If this conversation were produced in court tomorrow, would we still be comfortable with what we shared? #ArtificialIntelligence #LegalTech #AttorneyClientPrivilege #Confidentiality #AIGovernance #DataGovernance #TechnologyLaw #CorporateLaw #RiskManagement #FirstClause

Standardize the channel, not the tool. That one line resolves most of the AI confidentiality question in legal practice, and it is close to the opposite of what most firm policies do. What governs your client's confidence is not which model you typed into. It is which terms of use you were operating under - the channel, the tier, the model. The consumer and business versions of the same product are often governed very differently, and most lawyers cannot say which one they used. This is not the alarm genre. Retention by a vendor bound to confidentiality is not the same as waiving privilege - you entrust confidences to cloud mail and e-discovery every day. The duty is narrower, and it is competence: knowing which row of the terms you are standing in, and matching it to the sensitivity of the matter. I put the whole check on two pages, aimed at IP, technology, and in-house counsel. Five questions to ask before client material touches an AI tool. Two tables showing where the line falls. Free. No course, no upsell. If you cannot answer those five questions for a given task, you do not have an AI workflow. You have a confidentiality gamble. Link in the comments. Educational only, not legal advice. #LegalTech #LegalEthics #IPLaw #ArtificialIntelligence