Businesses are increasingly using AI to create logos, brand names, slogans and marketing materials. AI can generate branding ideas in seconds, but its use also raises questions about existing intellectual property rights.
What happens if an AI tool generates a logo or brand name that is identical or similar to an existing trade mark? And does it matter if the business never intended to copy another brand?
The short answer is that AI does not have to intend to infringe a trade mark for its output to create legal risk. However, whether trade mark infringement has actually occurred will depend on the circumstances, including how the sign is used and whether the relevant statutory test is satisfied.
Can AI itself infringe a trade mark?
Under current UK law, AI does not have separate legal personality and cannot itself be liable for trade mark infringement.
The UK Government has previously considered this specific issue. Its consultation on AI and intellectual property concluded that AI should be regarded as a tool or medium through which infringement may occur, rather than as the infringer itself. Responsibility remains with a legal person or organisation, with the courts determining liability according to the circumstances of each case.
This means that the important question is not whether AI has infringed a trade mark, but whether a person or business has used a sign in circumstances which amount to infringement.
For businesses using AI, this distinction is important. The fact that a machine generated the relevant material does not automatically remove the legal risks associated with using it.
Does intention matter?
A common misconception is that trade mark infringement requires deliberate copying. That is not generally the case.
The Trade Marks Act 1994 sets out a number of circumstances in which the use of a sign can amount to infringement. Depending on the circumstances, these include using an identical sign for identical goods or services, using a similar sign where there is a likelihood of confusion, or using a sign in relation to a trade mark with a reputation where the statutory requirements concerning unfair advantage, detriment or other prohibited use are met.
The focus is therefore on the use of the sign and the relevant legal test, rather than simply whether the person using it intended to copy another business.
For example, a company might ask an AI tool to create a logo for a new product. The prompt does not mention any competitor and the company has no intention of imitating another brand.
The AI nevertheless produces a logo which is similar to an existing registered trade mark.
If the business subsequently uses that logo commercially and the relevant requirements for infringement are satisfied, the absence of an intention to copy will not necessarily prevent a claim.
That does not mean that intention or knowledge is irrelevant in every respect. The circumstances in which the sign came to be used may be relevant to issues such as available defences, remedies and the overall assessment of the dispute.
What does Getty Images v Stability AI tell us?
The recent High Court decision in Getty Images (US) Inc & Ors v Stability AI Ltd [2025] EWHC 2863 (Ch) provides an important example of how established trade mark principles can apply to generative AI.
Getty Images brought proceedings against Stability AI concerning its Stable Diffusion generative AI model. Among other claims, Getty alleged that AI-generated images could reproduce Getty and iStock watermarks and that this amounted to trade mark infringement.
The Court found trade mark infringement in limited circumstances involving particular versions of Stable Diffusion and particular outputs. The Court's findings were fact-specific and did not establish that AI-generated content bearing a resemblance to a trade mark will automatically constitute infringement.
This is an important distinction.
The mere fact that an AI-generated image contains something that looks like a trade mark does not, by itself, answer the question of infringement. The Court still needs to consider matters such as the nature of the sign, the circumstances in which it appears, the relevant goods or services and whether the statutory requirements for infringement have been met.
The case therefore demonstrates that AI does not sit outside traditional trade mark law. Instead, existing legal principles have to be applied to a new technological context.
What if the AI generated the sign without being asked to?
This is where AI creates a particularly interesting legal problem.
A business may never have instructed an AI system to copy an existing trade mark. The similarity may have arisen entirely from the AI system's output.
The important questions are then likely to include:
- What exactly did the AI generate?
- Is the sign identical or similar to an existing trade mark?
- What goods or services is it being used for?
- Is the sign being used in the course of trade?
- Is there a likelihood of confusion?
- Who actually used the sign commercially?
- What did the business do after the AI generated it?
The distinction between generation and commercial use is particularly important.
An AI system may generate thousands of potential names or logos without any of them ever being used commercially. A trade mark dispute may therefore arise at a very different stage from the point at which an AI system simply produces an output.
Once a business adopts the output as its branding, however, the legal position can change significantly.
Who could be responsible?
Although AI itself cannot currently be liable for trade mark infringement, identifying the responsible party may not always be straightforward.
Depending on the circumstances, potentially relevant parties could include:
- the business using the AI-generated content;
- the provider of the AI system;
- the developer of the underlying model;
- organisations involved in deploying the technology; or
- other businesses involved in commercialising the output.
The UK Government has previously recognised these difficulties. In its consultation on AI and trade marks, it noted suggestions that responsibility could potentially depend on the circumstances, including the respective roles of an AI operator, user, provider and other parties. The Government concluded that existing trade mark legislation was capable of addressing these issues and that AI should be treated as a tool deployed under human direction.
The precise position will therefore depend heavily on the facts.
For businesses, the key point is that AI should not be treated as a legal "black box". Human involvement and responsibility remain central to the current legal framework.
What can businesses do to reduce the risk?
Businesses using AI to develop branding should consider the following safeguards.
1. Carry out trade mark searches
Before adopting an AI-generated name, logo or slogan, carry out appropriate searches for existing registered and unregistered trade mark rights.
2. Do not rely on the AI tool's assurances
An AI system may describe a brand name as "unique" or "original", but this is not a substitute for a proper trade mark clearance search.
3. Review AI-generated outputs before using them
AI-generated material should be subject to appropriate human review before it is placed on products, websites, packaging or advertising.
4. Keep an audit trail
Businesses should consider retaining records of how important branding was generated, reviewed and approved. This may become relevant if questions later arise about the origin or use of a particular sign.
5. Check the AI provider's terms
The business should understand the provider's terms concerning ownership, permitted use and responsibility for AI-generated outputs.
6. Consider trade mark registration
Where a business has developed a new brand, appropriate trade mark protection can help establish and protect its rights and provide greater certainty as the brand develops.
Final thoughts
AI can generate a brand name, logo or slogan in seconds. But it cannot guarantee that the result is free from third-party intellectual property rights.
The important question is therefore not simply whether AI intended to infringe a trade mark. AI does not have intention in the legal sense.
Instead, businesses need to consider what has been generated, how it is being used, whether the statutory requirements for infringement are satisfied and who is responsible for that use.
The Getty Images v Stability AI decision demonstrates that established trade mark principles can apply to AI-generated material, while also showing that infringement will depend on the particular facts.
For businesses using AI to develop new branding, the practical lesson is straightforward: treat AI-generated branding as a starting point, not a clearance exercise.
Appropriate trade mark searches and legal review before commercial launch can help identify potential conflicts before they become costly disputes.
How To Get In Contact
To find out more or if you require assistance with these matters, speak with our Intellectual Property Team on +44 (0)204 600 9907 or email info@culbertellis.com.
Accurate at the time of writing. This information is provided for general information purposes only and should not be relied upon as legal advice.
