This Authentix article has been published as a part of the Asia Video Industry Association’s Asia Video Report 2025. To learn more, please visit their website.
By Bharat Kapoor, Vice President, Online Brand Protection, Authentix
The creation of AI-generated visual content has profound implications for IP owners, especially in the context of copyright. When AI engines scrape images and videos to create new works, the critical question is whether such processing amounts to “reproduction” under copyright law. There also arises the vexed question of whether an AI-generated visual work is independently entitled to copyright protection. The few disputes that have so far reached courts have left many questions and concerns unaddressed. For example, in Getty Images v Stability AI (2023), Getty Images claimed before the UK High Court that Stability AI scraped millions of its copyrighted images to train its AI engine. While the court allowed the case to proceed to trial (finding that it involved a novel and previously undetermined question of law) it remains to be seen whether the parties decide to settle, thus scuttling the development of judicial guidelines on the subject. In Kneschke v LAION (2024), the Hamburg Regional Court in Germany rejected a similar claim by a stock image photographer against LAION, a non-profit AI company. Somewhat controversially, the court allowed LAION the benefit of an exemption normally reserved for research organizations. If courts begin to lean in this direction globally, copyright owners will struggle to be renumerated fairly. Given the deep links between certain non-profits and major tech companies, such an outcome might even be seen as patently unjust. Meanwhile, in Thaler v. Perlmutter (2023), the US District Court in Columbia upheld the US Copyright Office’s denial of copyright in an image created using AI, ruling that “human authorship” is a necessary ingredient of copyright ownership. But the ruling arguably fails to provide guidance in situations involving human authorship assisted by AI, or the highly creative use of an AI engine where the user goes beyond dictating mundane commands.
Alongside copyright, an overlooked area involves the use of trademarks in the creation of AI-generated visual content. To illustrate, Authentix Online Brand Protection conducted an experiment where multiple AI apps were instructed to create hundreds of images using well-known trademarks in (arguably) dilutionary contexts. The results were mixed. Some engines did not use the specified trademarks, contrary to our commands. Some slightly altered the specified trademarks and then used these versions. And some used all kinds of trademarks uninhibitedly, in contexts that would probably displease trademark owners (such as Ronald McDonald advertising KFC). Depending on context and jurisdiction, such uses might be prohibited as dilution/tarnishment, or permitted under fair use/criticism/parody exceptions.
In the end, AI can make the process of creating high-quality visual content quicker, easier, cheaper, more accessible and more scalable. This can benefit not only novice or low-budget content creators, but also large corporations wishing to expand their reach through social media. However, IP rights owners will likely argue in favor of laws and regulations governing the use of IP by AI engines. Rights owners may also advocate the adoption of licensing regimes (as a possible harbinger, OpenAI has reportedly been offering news organizations between USD 1 to 5 million in licensing fees to train its AI model). In both cases, copyright-related principles will probably be easier to concretize into laws and norms than trademark principles, and also receive higher prioritization. Yet, in all this, a fact that could be missed is that Western nations no longer monopolize the development of AI. AI engines that place IP owners at risk may well emanate from emerging Asian economies, with contrary viewpoints on the boundaries of IP protection. The bigger challenge for IP owners, therefore, may lie in ensuring the global acceptability of IP norms and the consent of governments and AI developers from emerging economies in the framing of such norms.
About Authentix Online Brand Protection
Authentix provides some of the world’s most recognizable brands with sophisticated online brand protection tools and services to address a broad range of online infringement and counterfeit risks. From global online surveillance and enforcement, online investigations and site takedowns, target verification, and even offline investigations, Authentix helps major brands proactively reduce the threat of counterfeit products being sold online. Our online brand protection solutions combine cutting-edge technological tools and expert analysis to reduce infringements for our clients on online marketplaces, social media platforms, and websites by up to 90%.
Want to see what Authentix OBP can do for your company? Our team is committed to working with you, understanding your unique challenges and requirements, and providing tailored solutions that best fit your industry and the unique challenges your brand faces that make a tangible difference to your brand. Schedule a consultation with one of our brand protection experts to learn more.
Bharat Kapoor, Vice President, Online Brand Protection, Authentix
Formerly the CEO of Strategic IP Information Pte, now a part of Authentix, Bharat has been focused on building effective global brand protection programs for over 250 brands across sectors such as fashion, FMCG, pharma, cosmetics, sporting goods wines, and spirits. He has led the development of the best-in-class machine learning-based online brand protection technology and developed effective global teams to investigate counterfeiting and piracy on the ground in China, Asia, and Latin America. Bharat Kapoor has a B.A. Honors from the University of Technology, Sydney. Prior to joining SIPI, he was an Associate Director at Ernst & Young India, where he led the business development and advisory services program for companies operating in the media, e-commerce, and advertising sectors.