Consumers are increasingly relying on large language models (LLMs) to do their online shopping. While this change in consumer behaviour will have wide-ranging impacts, this article explores some considerations for how brand owners may protect and enforce their trademarks as this area develops.
AI shopping tools
There is no shortage of AI shopping options, and each tool has its own features and capabilities:
General-use LLMs such as Microsoft Copilot and ChatGPT are able to respond to shopping-related queries and provide recommendations. These are not shopping-specific tools so do not offer features such as real-time price tracking.
Amazon has launched a beta version of its AI chatbot Rufus. It is integrated into the Amazon website so it can recommend buying options from Amazon’s marketplace, compare product features, display real-time pricing, and give recommendations based on criteria such as ‘best value’.
OpenAI’s “agent” tool goes further and can complete tasks online for the user, including selecting a product, putting it in the shopper’s basket, and filling in relevant delivery information. It leaves it to the user to confirm the purchase and authorise payment so the use of sensitive data is avoided.
Perplexity’s shopping-specific features allow users to complete a purchase on its own platform (for selected products and merchants), without ever going directly to the brand’s website.
Enforcing against infringements
With these developments come important considerations for brand owners. One concern is that LLMs could recommend counterfeit or otherwise infringing products where they fit a shopper’s criteria, particularly in cases where a user may have specified that they are shopping on a budget. Although the average shopper may be wary of counterfeits when searching e-commerce sites themselves, their attention may be lower than usual where they are relying on AI in the context of trusted e-commerce sites. The risk would be exacerbated in the case of lower-value purchases, and where shoppers use more involved AI tools.
It is therefore more important than ever that brands monitor online infringements closely and take steps to have counterfeit or otherwise infringing listings removed. Most major e-commerce platforms have intellectual property infringement policies in place, with tried and tested complaint procedures for brand owners to use, including eBay’s Verified Rights Owner (VeRO) programme. As always, brand owners should be careful not to make unjustified threats against secondary infringers that use e-commerce sites to retail third-party products.
Evidence in trademark proceedings
AI shopping bots are also likely to have an impact on the evidence that brand owners produce to prove their reputation and goodwill in trademark proceedings. Several start-up companies offer the ability to monitor a brand’s presence in AI chatbot results. This gives businesses valuable knowledge as to which AI tools give their brand the greatest exposure compared with other brands; if a brand is cited in shopping-related queries fewer times by one LLM than another, brand owners can take steps to remedy that.
These statistics could be strong evidence when attempting to prove that a trademark has a reputation and/or goodwill in the context of tribunal or court proceedings. Other types of evidence, such as the number of visitors to a brand’s website, may become less relevant as consumers rely more on AI shopping tools and less frequently visit brand owners’ websites directly – as mentioned above, some AI tools offer their own platforms to complete online shopping without the customer ever visiting the merchant’s website directly.
Longer-term questions
There will be further considerations for brand owners as this area develops, including whether to shift their advertising efforts towards AI tools, rather than direct to consumers, and the possibility that product brands could have reduced importance online over time.