Last year, we discussed the benefits and potential dangers of artificial intelligence in “Mitigating the Risks of AI in your Business.” This blog goes into a deeper dive on how intellectual property concerns related to AI are now showing up in courtrooms, and how businesses can best protect their brand and marketplace identity while still enjoying the benefits of AI systems.
Generative AI: Intellectual Property Risks
Many AI systems are trained on large datasets containing licensed material, public-domain works, proprietary data, and other content pulled from the internet. This training process opens the door to potential liability issues for AI companies and their users. Legal concerns arise due to the material the systems use, what appears in the resulting output, and how people and businesses ultimately use that output. Specifically, intellectual property disputes for AI companies and the businesses that use them generally arise at two points:
- Infringement: Intellectual property infringement could occur when the AI’s output looks or behaves close enough to an existing brand or protected work.
- Invention: Sometimes called an “AI hallucination,” this occurs when an AI system invents false information and attaches it to a real, recognizable trademark or brand name.
Modern Cases: AI Liability
When an AI tool generates a logo, slogan, or other content confusingly similar to someone else’s protected work or brand, using that output can create the same intellectual property concerns as human-created content. A recent example is Advance Local Media LLC v. Cohere Inc., in which a group of major news publishers sued the AI company Cohere, alleging its AI systems produced outputs containing reproductions of their copyrighted works and, in some instances, falsely attributed “AI hallucinated content” to the publishers’ brand. In 2025, the court denied Cohere’s motion to dismiss several of the publishers’ claims, allowing copyright infringement, secondary liability, and trademark claims to proceed. Although the case has not yet determined Cohere’s ultimate liability, it demonstrates that information generated through AI systems are subject to the same intellectual property laws that regulate human-created content. It also highlights a risk for businesses: AI-generated content may reproduce your protected works or incorrectly suggest that your brand created or endorsed AI-generated information that it did not. This case also cautions AI users against trusting AI-generated content being put forth as facts endorsed by external entities.
A related dispute is unfolding in Andersen v. Stability AI Ltd., where visual artists sued several AI companies over the alleged use of their copyrighted artwork by AI-image generation systems. Here again, the decision was not a final ruling, but claims of copyright infringement were found plausible enough to proceed to discovery. This case illustrates that AI companies will be scrutinized, and perhaps held accountable, when using others’ content to train their systems. For businesses using AI-generated images or other creative content, cases like Andersen underscore the importance of understanding who bears the risk if an output triggers an infringement claim, which should be addressed in the AI company’s contracts.
The Bottom Line for Your Business
The use of AI can boost business efficiency, but using it safely requires more than just proofreading the product. It is imperative to read the contract of an AI producer before putting information into its system. AI platforms have their own terms of use, and those terms may also differ between free, individual, or enterprise versions of the same AI system. Before adopting an AI tool, businesses should pay particular attention to:
- Ownership of inputs and outputs. Review the ownership rights the AI provider claims in information submitted to its system and make sure you understand who owns both the information submitted into the system, and the outputs that are generated.
- Training and data use rights. Determine whether the system provider may use prompts, uploaded documents, or other data to train or improve its basic models which are used for other clients. For businesses handling confidential information, the right to opt out of training, or a contractual prohibition on such use, could be particularly important.
- Confidentiality and data retention. Review how long the provider may retain information, where it is stored, and whether it may be shared with sub-processors or other third parties.
- Intellectual property risks. If an AI-generated image, text, or other output leads to an infringement claim, determine who bears the cost of defending that claim. Review whether the provider offers indemnification, what types of claims it covers, and what exclusions or conditions limit the protection.
- Changes to the service. AI systems develop quickly. It is important to be aware whether and when the AI provider may change its model, data practices, terms, or other features in the contract, and whether the user will receive advance notice or have termination rights.
Reviewing the contract of your AI system is only one part of protecting your business. Companies should also establish internal rules governing how employees use AI. An effective AI policy can identify approved platforms, prohibit or restrict the submission of proprietary or confidential information, require human review before generated material is used, and establish additional review for higher-risk uses, such as branded content.
Businesses should also protect the other side of the equation: their own brand. This may include registering important trademarks and copyrights where appropriate, maintaining records of human-created work, and monitoring the marketplace for confusingly similar or unauthorized uses of company names, logos, designs, and other protected materials.
Ultimately, while AI can be a valuable business tool, it should be treated as a starting point, not a substitute for human judgment. Businesses that pair AI’s efficiency with human review, clear internal policies, and thoughtful protection of their intellectual property, can take advantage of what this technology has to offer, without unnecessarily putting their business at risk.
As AI technology continues to evolve, a proactive trademark strategy is a business imperative. If you’re unsure where to start or are concerned about how to protect your business, reach out to our experienced business attorneys at Hamelburg Law, LLC.
This blog was written by Jamie Kent Hamelburg, an attorney at Hamelburg Law, LLC, and Emily Roberts, an intern who is currently a student at the University of Michigan.
Note: The content in this Blog is for informational purposes only and should not be acted upon without first consulting legal counsel. It is not intended to constitute legal advice.
