Intellectual Property Consulting

In a MAJOR ruling for European copyright law, the Munich Regional Court has sided with Germany’s music rights society GEMA against OpenAI, finding that the company’s ChatGPT model unlawfully used copyrighted song lyrics in its training and responses. The decision, issued this morning, marks the first major European court judgment holding an AI company liable for using protected works without a licence. I got into AI through being Director of Legal Affairs and Regulatory Compliance in IMRO, the Irish counterpart of GEMA - and I know the people in GEMA - so this is very interesting to me. The case centred on GEMA’s allegation that OpenAI trained ChatGPT on its repertoire of German song lyrics, allowing the chatbot to reproduce works by artists such as Helene Fischer and Herbert Grönemeyer. The court agreed, concluding that the model’s ability to reproduce lyrics word for word demonstrated that the works had been used in training. It ruled that OpenAI is liable for copyright infringement and prohibited ChatGPT from reproducing lyrics from GEMA-represented artists unless a licence is obtained. The court also held that the European Union’s Text and Data Mining exceptions cannot shield generative AI systems that “memorise” and reproduce copyrighted material. This reasoning undermines one of the primary legal defences AI developers have relied upon in Europe. While damages will be determined in a separate proceeding, the court’s finding of liability alone sets a powerful precedent. OpenAI has announced plans to appeal. The 42nd Civil Chamber of the Munich Regional Court had indicated its position in September, when it observed that the model’s outputs could not be explained without training on copyrighted material. The final judgment confirmed that assessment. For the wider AI sector, the ruling suggests that AI companies operating in the European Union may need explicit licences for any copyrighted content used in model training or risk litigation. The decision also has regulatory implications. It aligns with growing momentum within the EU to enforce transparency and rights-holder protections under the AI Act and the Copyright in the Digital Single Market Directive. The GEMA v OpenAI ruling diverges sharply from Bartz v Anthropic in the United States. In Bartz, Judge Alsup found that AI training on copyrighted material could qualify as fair use, meaning no licence is required when the use is deemed transformative and non-substitutive. He viewed training as an analytical process that teaches the model general patterns rather than reproducing expression. The Munich court took the opposite view, holding that using protected works in AI training without permission constitutes reproduction requiring a licence. This illustrates the growing divide between the U.S. model, where fair use can exempt AI developers from licensing duties, and the European approach, which treats copyright as an enforceable economic right demanding prior authorisation.

Branding without trademarking is like building a house on land you don't own. When I launched Business Class, the first thing I did wasn't design the curriculum — it was lock down the trademark. Here's what most founders get wrong about intellectual property protection: 🎯 Trademarks are context-specific: You don't trademark a name "in general." You trademark it by class: → Business Class is in Class 41 (education/digital courses) → If I launch merch? That's Class 25 (apparel) → mobile app? Class 9 (software) This is why two companies can have the same name and coexist — if they're in different classes and not causing market confusion. 🔍 How to research your mark: Use the USPTO TESS database to check availability by class, not just globally. Someone might own "Business Class" for travel booking, but that doesn't automatically conflict with my education business. Context matters. 📝 Two types you need: - Word mark: Protects the name regardless of how it looks - Design mark: Protects your logo/visual identity You want both. ⚠️ Where most applications eie: - Similarity isn't just spelling. - The USPTO evaluates: → How it sounds → How it looks → Whether average consumers could confuse the two "Confusingly similar" kills more trademark applications than anything else. 🛡️ Use it or lose it: Once it's yours, defend it aggressively. If your mark becomes generic (like Aspirin did for Bayer), you lose exclusive rights. A trademark isn't permanent if you treat it as optional. The bottom line: IP is leverage. If you're building a brand you want to scale, protect the asset before you promote it. Because if you don't — someone else will. For founders in the comments: What's the biggest IP mistake you've seen (or made)? Join my Substack for more wisdom & war stories: sophiaamoruso.substack.com

The US Copyright Office has just released its Part 3 Report on Generative AI Training, and it addresses the elephant in the dataset: Can AI companies use copyrighted content to train their models without permission or payment? The report says this is not a grey area. Training on copyrighted works is not automatically protected under fair use, particularly when conducted at scale and for commercial use. The report outlines multiple stages that can raise infringement claims from scraping and dataset curation to model training and the generation of outputs. The Office explicitly rejects the idea that “publicly available” content online is free for use in AI training. That position, often relied on by developers, does not hold up under copyright scrutiny. The fair use analysis is direct: 𝐏𝐮𝐫𝐩𝐨𝐬𝐞: The use is commercial, high-volume, and systemic, not limited or research-driven. 𝐀𝐦𝐨𝐮𝐧𝐭 𝐮𝐬𝐞𝐝: Full works and large repositories are routinely copied. 𝐌𝐚𝐫𝐤𝐞𝐭 𝐢𝐦𝐩𝐚𝐜𝐭: AI outputs often compete with the original works and may displace licensed content. 𝐓𝐫𝐚𝐧𝐬𝐟𝐨𝐫𝐦𝐚𝐭𝐢𝐯𝐞𝐧𝐞𝐬𝐬: Using expressive content to generate similar expressive content is unlikely to qualify. The Office states: 𝘛𝘩𝘦 𝘤𝘰𝘱𝘺𝘪𝘯𝘨 𝘪𝘯𝘷𝘰𝘭𝘷𝘦𝘥 𝘪𝘯 𝘈𝘐 𝘵𝘳𝘢𝘪𝘯𝘪𝘯𝘨 𝘵𝘩𝘳𝘦𝘢𝘵𝘦𝘯𝘴 𝘴𝘪𝘨𝘯𝘪𝘧𝘪𝘤𝘢𝘯𝘵 𝘱𝘰𝘵𝘦𝘯𝘵𝘪𝘢𝘭 𝘩𝘢𝘳𝘮 𝘵𝘰 𝘵𝘩𝘦 𝘮𝘢𝘳𝘬𝘦𝘵 𝘧𝘰𝘳 𝘰𝘳 𝘷𝘢𝘭𝘶𝘦 𝘰𝘧 𝘤𝘰𝘱𝘺𝘳𝘪𝘨𝘩𝘵𝘦𝘥 𝘸𝘰𝘳𝘬𝘴. This is a key clarification for the industry. Developers relying on generic fair use claims will have to prove that their specific training methods and outputs meet the legal threshold but most won’t. The report also addresses and rejects common defenses: 📌AI training is not a “non-expressive” use. 📌Public access is not the same as permission. 📌Training on infringing datasets attracts stricter scrutiny. While the report stops short of policy prescriptions, it identifies extended collective licensing as a possible solution where voluntary markets fall short. It also notes legal and operational barriers that would need to be addressed for such a system to work. The report can be accessed at: https://lnkd.in/gD8fn-jA #copyright

🚨AI, IP Infringement & the Legal Gaps ............. Who Owns the Blame? 🔎 Rise of AI-Driven IP Infringement AI is now capable of generating art, music, literature, and even inventions. This has blurred the lines between creator, owner, and violator. The big question: When AI infringes, who is accountable : the developer, the user, or the AI itself? ⚖️ International Perspectives US & EU frameworks lean on traditional copyright and patent systems. AI-generated works are generally not recognized as authorship. Liability often falls back on the human programmer or deploying entity. Global legal systems are struggling to move beyond human-centric laws. 🇮🇳 The Indian Context: Laws & Loopholes Indian Copyright Act (1957) defines “author” as a human, leaving no clarity for AI outputs. Courts have not yet adjudicated directly on AI authorship or liability. Patent law loopholes: AI-generated inventions face rejection as inventors must be human. Lack of clear policy guidance puts businesses and creators at risk. 🚧 Key Challenges Identified Attribution Gap: No legal clarity on who gets credit. Accountability Gap: Ambiguity over liability in AI-driven infringements. Regulatory Lag: AI evolves faster than lawmaking. Cross-Border Conflicts: Different jurisdictions interpret AI rights differently. ✅ Recommendations proposed by Experts and legal luminaries Update Copyright & Patent definitions to account for AI. Consider a “shared liability model” between developer, deployer, and user. Introduce AI-specific IP guidelines for both ownership and infringement. Strengthen international cooperation for harmonized AI-IP laws. 📚 Sources Referenced SCC Online Blog: Legal Accountability for AI-Driven Intellectual Property Infringements – An Analysis of International and Indian Laws (2025) Comparative insights from US, EU, and Indian IP law frameworks. #Corporategovernance #Independentdirectors #IPR #AI #Copyrights #Infringement

Your IP strategy that worked in the US won't work in India. This isn't a minor tweak. It's a complete rethink. Every country builds its IP laws around its own economic priorities, industries, and policy goals. What's patentable in Europe or the US can be flatly rejected in India. Businesses find this out the hard way, after filing. India's patent law has stricter criteria than most Western systems. An innovation that clears the bar abroad may still fail scrutiny here if it doesn't meet India's specific standards for novelty and inventive step. The Indian Patent Office processed over 68,000 applications in 2025 alone. And in 2026, new patentability guidelines were introduced. The rules you researched last year may already need a second look. Here's what businesses entering India often overlook: → Trademark classes and filing conventions differ from WIPO norms → Prior disclosure rules can invalidate a patent if timelines aren't managed carefully → Enforcement mechanisms work differently. Registering your IP is only half the job The good news: India is actively refining its framework. Processing times have already dropped 20% following recent amendments. The system is getting sharper — but so must your strategy. Copying a global IP playbook into the Indian market is one of the most expensive assumptions a business can make. Have you had to rethink your IP approach specifically for India — or are you still working off a global template?

The data from the Mitsui & Co. Global Strategic Studies Institute presents a subtle but decisive shift in how #semiconductor leadership is being secured—not just through manufacturing scale, but through where knowledge is legally anchored. At a surface level, the numbers are straightforward: Most global leaders—Tokyo Electron, Samsung Electronics, Applied Materials, TSMC, and ASML—file a significant share of patents in the United States, often exceeding 70–90%. In contrast, Chinese entities like Chinese Academy of Sciences and NAURA Technology Group file almost entirely within China (~98%). But the strategic signal sits beneath this distribution. First insight: The US remains the global enforcement ground for IP. Filing in the US is not just about market access—it is about legal strength. The US patent system still acts as the most credible arena for defending high-value semiconductor innovations. This explains why even non-US companies anchor their IP there. Control in semiconductors is as much about litigation readiness as it is about fabrication capacity. Second insight: China is building a self-contained innovation loop. The near-total domestic filing by Chinese institutions signals a deliberate inward strategy. This is not a lag—it is a design choice. By concentrating patents locally, China is strengthening internal supply chains, reducing external dependency, and creating a protected innovation environment aligned with national priorities. Third insight: Two parallel IP ecosystems are forming. One is globally integrated, anchored around the US system. The other is domestically reinforced within China. Over time, this divergence could lead to limited interoperability—not just in technology standards, but in legal enforceability of innovation. Fourth insight: Patents are becoming strategic assets, not just legal instruments. In semiconductors, patents define control over process nodes, materials, lithography techniques, and equipment precision. Owning patents in the right jurisdiction determines who captures long-term economic value, who sets pricing power, and who controls ecosystem dependencies. This is where the conversation shifts from “innovation” to “ownership of innovation outcomes.” manufacturing builds the factory, but patents own the blueprint of the factory. One scales output, the other governs who is allowed to scale. For leadership teams, this has clear implications: R&D without a jurisdiction strategy is incomplete Market expansion must align with IP protection zones Partnerships need to account for where knowledge will be legally held National policy and corporate strategy are now tightly interlinked in deep tech sectors The semiconductor race is no longer only about nanometers. It is about where ideas are registered, defended, and monetized. Those who understand this will not just build technology—they will control its future value. DC* Dinwins

Patents are not just legal shields, they’re powerful business assets. The way you license them can define whether your invention remains locked in a drawer or becomes a global phenomenon. Effective Patent Licensing Strategies include: 📌Exclusive Licensing – Giving one licensee full rights, often in exchange for higher royalties. 📌Non-Exclusive Licensing – Allowing multiple licensees, maximizing reach and volume. 📌Cross-Licensing – Two or more companies exchange patent rights, reducing litigation risk and fostering innovation. 📌Field-of-Use Licensing – Granting rights limited to specific industries or applications. 📌Sublicensing – Allowing the licensee to grant rights further, useful for scaling. 📌Case in Point: IBM’s Patent Licensing Strategy IBM is a textbook example. In the 1990s, IBM shifted focus from just making hardware to actively monetizing its vast patent portfolio. By licensing patents across industries, semiconductors, software, and IT, IBM generated over $1 billion annually in licensing revenue. 📌This strategy not only diversified revenue streams but also positioned IBM as a powerhouse of innovation while avoiding unnecessary litigation. A well-structured licensing strategy can transform patents into sustainable revenue, open new markets, and build strategic partnerships. #PatentStrategy #Innovation #Licensing #IPR #TechnologyBusiness #Patents

Your customer purchased 10,000 licenses for your product over a 3-year term, then laid off 2,000 people, now they’ve got 2,000 empty seats on the shelf. They may not be worried about their shelfware, but your customer success team should be. Proactively getting those licenses reactivated is one of the most important risk mitigation tactics your CS team can embrace, and they’ve got to be the ones to do it. Sales has little incentive to jump in, as it doesn’t impact their quota, and their focus is likely on acquiring new business, not maximizing value from existing ones. Simply put, they aren’t compensated to do it. So, what’s your best strategy here? Effective multithreading. When you’ve established relationships with enough people within an organization, and you’ve delivered your value story across the board, they should realize how your product is helping their company increase revenue. In doing so, you’ve significantly broadened its use case and necessity. You’ve got proven results. Talk to the network of decision makers you’ve built, talk to those who are influential, and find new champions. By educating them on the cost-saving opportunities of reactivating unused licenses and the potential for increased productivity and efficiency, you can create internal buy-in. This is how you ensure customers get the most value from their investment, while maximizing license utilization.

Big victory for #copyright (holders) in GEMA vs. OpenAI. Today, the Munich Regional Court issued Europe’s first judgment finding that OpenAI infringed copyright by (i) memorising (song) texts and (ii) displaying parts thereof in the output. The judgment on behalf of music collecting society GEMA is groundbreaking on many topics: “Both the memorisation in the language models and the reproduction of the song lyrics in the chatbot's outputs constitute infringements of copyright; not covered by any limitations, in particular [..] text and data mining.’ "Memorisation [...] occurs when the language models not only extract information from the training data set during training, but also completely adopt the training data in the parameters specified after training. [...] The song lyrics at issue are reproducibly defined in the models.”   "If not only information is extracted from training data during training, but works are also reproduced, this does not constitute text and data mining [..] In the case of reproductions in the model [..] the exploitation of the work is permanently impaired and the legitimate interests of the rights holders are thereby infringed.” “[T]he defendants also unlawfully reproduced and made publicly available the song lyrics in question by reproducing the lyrics in the chatbot's outputs. The original elements of the song lyrics would always be recognisable in the outputs.” “The defendants, and not the users, are responsible.” The judgment, if upheld, will have far-reaching consequences – a very welcome clarification on copyright law in the era of GenAI. Big congrats to GEMA, Robert Heine and his team for bringing this important precedent! Unofficial courtesy translation of Court’s press release attached.