The AI & Legal Update #18 - 🥊 Mickey Mouse on the Prowl and Pending AI Regulation Ban

Hi friends! The OG of suing tech companies for copyright issues - Disney - is back. The mouse filed suit against AI video and image maker MidJourney for operating "a virtual vending machine, generating endless unauthorized copies" of Disney's iconic works. Burn. And there's more. Read on for AI regulations, big tech moves, and more AI in court. Even a new proposed Federal Rule of Evidence. Woah.

📘 Read time: 7 min, 28 sec 

AI & Regulatory

⚖️ Europe’s regulators issued their first big fines under the new Digital Markets Act, slapping Apple with €500 M and Meta with €200 M. Trump called the regulation “economic extortion” in February but that doesn't change the compliance requirements.  Here’s GC AI’s checklist for complying with the Digital Markets Act. Tl;dr - it applies to "gatekeepers", i.e. companies that have >$7.5B annual revenue in Europe the last three years. Speaking of, this FT article about Zuckerberg's transition to "MAGA Mark" is good. 

🦺 A 2025 global study from KPMG finds most Americans are concerned about AI risks, with 70% of respondents believing AI needs more regulation. Unsurprisingly, people are more willing to use AI for creative tasks than for high-stakes decisions like medical diagnoses or hiring. What about law? Well, you know how we feel. Use an AI that gives you verbatim quotes and the ability to check citations. Like, ahem, GC AI. 

🧩 The National Conference of State Legislatures is tracking AI legislation across all states heading into 2025. Related, in May, the US House of Representatives passed a provision in the "One Big Beautiful Bill Act" (OBBBA) that would impose a 10-year moratorium on state and local enforcement of "any law or regulation regulating artificial intelligence models, artificial intelligence systems, or automated decision systems." If enacted, the moratorium would 

• Preempt existing state AI laws in California, Colorado, New York, Illinois, and Utah

• Impact the over 1,000 pending AI bills across state legislatures 

• Affect regulatory oversight across finance, insurance, education, healthcare, and other sectors

  
  GC AI's prediction? It won't pass. We'll keep you posted! 

Big Tech

🔨 Google had a major antitrust setback last month when Judge Brinkema found Google liable for "willfully acquiring and maintaining monopoly power." See GC AI summary. Interestingly, what could save Google from having to divest parts of its ad business? Competition from OpenAI.

Related, OpenAI says it would buy Google's Chrome browser if a federal court orders it to be spun off. Fun fact: Google’s Gemini AI has 35 million daily users (350M monthly), much less than ChatGPT's 160M daily​. Egads. OpenAI also announced $10 billion in annual recurring revenue (ARR) as of June 2025. Your editor worked on a deal with them when OAI was effectively pre-revenue. They grow up fast! 

💰 Coincidentally, more money does bring more problems. OpenAI said in May it'll remain under nonprofit control while transitioning commercial operations into a Public Benefit Corporation. But .. they're working on evolving their deals with Microsoft so MSFT doesn't get as big a profit share and loses some control and other provisions MSFT has today. This is moving too fast for your editor to keep up other than to say, game on, and congrats to OpenAI legal, on supporting some of the most strategic maneuvering your editor has seen in years. PS - for those of us who manage comms, liking this non-statement from the companies: "talks are ongoing and we are optimistic we will continue to build together for years to come." Missing a comma though. 🫠

🧠 Speaking of big legal team kudos, per OpenAI CEO Sam Altman, MAGA Zuckerberg is recruiting the very best in AI and willing to pay big bucks for them. In a f̶a̶i̶r̶l̶y̶ ̶o̶b̶v̶i̶o̶u̶s̶ ̶r̶e̶g̶u̶l̶a̶t̶o̶r̶y̶ ̶s̶k̶i̶r̶t̶i̶n̶g̶ smartly-structured and creative deal, Meta paid $14B to invest in AI technology company Scale AI and provide return to Scale's investors and employees in order to hire Scale AI's now ex-CEO, former wunderkind Alexandr Wang, to join Meta. Strategists are saying Meta is scrambling. But people have discounted Zuck before and been wrong. He told advertisers recently that with AI-powered ad targeting, "at this point [] we are just better at finding the people who are going to resonate with your product than you are." Fair. 

🥭 AI has shuffled the Big Tech leaderboard: the old term, FAANG, is out, MANGO is in. That’s Meta, Apple, Nvidia, Google, and Open AI.

AI & Legal

🤖 Meta plans to automate up to 90% of its privacy and societal risk reviews, replacing human assessors with AI, including for algorithm updates once debated by policy teams.

⚖️ Exec search firm Russell Reynolds polled GCs and found 75% of in house legal departments are either exploring or piloting AI tools, but only 28% of legal leaders say they are confident they have the requisite skills. Two notes from the editor: (1) the number is similar to what we see in our polls, with 72% of teams having an explicitly AI goal in 2025, and (2) if you feel you or your team don't have the requisite skills to pick up AI for legal work, by all means, do yourself a favor and take one of our classes.

Similar stats in EY's 2025 Law General Counsel study, which found that 75% of legal departments are refining their tech strategies - and 1 in 4 now prioritize GenAI. Regulatory compliance remains the top priority, yet 60% say staying ahead of regulatory change is still a challenge.

🧑‍⚖️ In April, in the civil case Ferlito v. Harbor Freight, Judge Brown of the E.D.N.Y. upheld an expert’s use of ChatGPT to support - but not form - his opinion. The court found no FRE 702 Daubert issues, marking a pragmatic stance: AI can assist, but expert judgment must lead. Read GC AI summary here. 

🧑‍⚖️ The US Committee on the Rules of Practice and Procedure proposed and sought public comment for new Federal Rule of Evidence 707 that would require disclosure of the use of AI for evidence generation.

AI & Corporate Adoption

🏷️ Pinterest will now automatically label AI-generated images and offer filters to exclude them.

👩‍🎓 Unemployment for recent college grads is 5.8%, compared to 4.2% overall - a record gap. Pandemic effects and declining college ROI may be factors, but The Atlantic suggests a deeper shift - AI could be replacing entry-level white-collar jobs that once served as career starting points. Law-school applications are surging again, echoing the post-2008 retreat to grad school. Was that any of you?  

Lots of studies debating the efficacy of AI. 

• Bain finds 95% of companies are using GenAI, 

• But McKinsey says only 1% of U.S. companies report they have scaled their AI investments effectively, with 43% still in pilot phases.

• The WSJ has the doomsday saying most companies report <10% cost savings and <5% revenue increase from AI. 

Your editors take? Adopt it anyway. The alternative of not using AI? Irrelevance. Gah.

AI Globally

🧑‍⚖️ The UK House of Lords is advocating for "risk-based" regulation and a principles-based approach with sector-specific rules rather than a comprehensive AI Act like the EU's. 

🏎️ Sam Altman told the Senate Commerce Committee that AI will be "at least as big as the internet, maybe bigger", emphasizing the need for infrastructure investment and warning that China is rapidly closing the AI gap with the U.S.

AI & Intellectual Property

🌐 OpenAI's Chris Lehane is pushing back on the "open source vs. closed AI" debate, saying we need both to keep the US ahead.

📄 The US Copyright Office released its third report on AI and copyright, recommending a “flexible approach” to training models on copyrighted materials. Two days later, President Trump fired the report’s author, Register of Copyrights Shira Perlmutter, following his dismissal of Librarian of Congress Carla Hayden. Big tech’s attendance at the State of the Union paid for itself? Separately, the Copyright Office also launched a new initiative to develop copyright registration guidance for AI-assisted works, aiming to clarify how much human input is needed for copyright protection. This comes after their previous determination that purely AI-generated content lacks human authorship and thus copyright protection.

🎬 AI assisted films are eligible for Oscars, and AI use will "neither help nor harm" nomination chances, says The Academy of Motion Picture Arts and Sciences. 

🎭 SAG-AFTRA has filed unfair labor practice charges against Epic Games over the use of AI to replicate James Earl Jones's iconic Darth Vader voice in the game Fortnite. The union alleges Epic violated federal labor law by using AI voice technology without proper agreements. Jones, who passed away in 2023 at age 92, had previously signed an agreement allowing his archival voice recordings to be used for future Lucasfilm projects, and his family specifically granted permission for the Fortnite usage. See GC AI’s analysis - with exact citations (!) below.

🪄 Disney and Universal have filed a copyright suit against Midjourney, accusing the AI image generator of copying protected characters. The suit marks the first time major studios have gone to court over AI-generated images. In other words, Hollywood would like its cut of the $300M Midjourney made last year. Disney’s GC Horacio Guitierrez (who your editor met at the amazing Madrona GC conference in March, thank you Joanna Black) put it plainly: “Piracy is piracy, and the fact that it’s done by an AI company does not make it any less infringing.” See images from Midjourney below. Who are the characters? Can you tell? 🤪

🤝 In April, The Washington Post entered a "strategic partnership" with OpenAI that allows ChatGPT to display summaries, quotes, and direct links to articles, but “always with clear attribution.”  Here’s a GC AI breakdown of what this means for OpenAI’s fair use defense in it case against The New York Times.

News, Notable, & Random

✝️ New Pope Leo XIV named AI one of the “main challenges for humanity” while laying out his vision for the papacy. 

💼 AI errors are now insurable! Lloyd's of London is offering coverage for AI hallucinations and chatbot errors - like Air Canada’s chatbot promising a bereavement discount the company didn’t offer, or Virgin Airlines’ bot, which scolded a customer for asking about “Virgin Money” - their own trademark.

Events & AI

🏡 New on In-House Connect: our webinar on AI prompting in Word for In-House Legal. Discover how AI can help streamline redlines, comments, issues lists, and more. AI newbies or advanced users are all welcome.

🎥 Check out the panel we did with Doug Mandell, Marie Ma, and Frances Pomposo discussing how AI can amp up your legal career. Hosted by Anne Kerwin, the in-house specialist. 

Upcoming Classes

👩🏻‍🏫 AI Prompting for In-House Legal - 101 on Thur, 6/26 at 9 am PT / 12 pm ET

👩🏻‍🏫 AI Prompting for In-House Legal - 101 on Tue, 7/8 at 9 am PT / 12 pm ET

👩🏻‍🏫 Advanced AI Prompting for In-House Legal - 201 on Wed, 7/16 at 10 am PT / 1 pm ET

Sign up here

How did you like this newsletter? What can we do better? Click to send feedback. 

❤️  I loved it!

😐 Good but could be better. 

👎 It was meh.