A long-standing trend of the tech world is “industry disruption,” which often takes the form of taking action first and dealing with legal consequences later, whether that is new regulations to contain Uber and Lyft or judicial rulings reigning in trademark infringement.
While a number of tech company actions have inspired laws and guidelines to mold to new ways of doing things, some regulations are foundational enough to evade disruption. This seems to apply to a recent trademark infringement lawsuit arising between Google and Gemini Data over the use of the name “Gemini” for an AI tool.
It is hard to see how Google will convince the Northern California District Court that this is not trademark infringement, plain and simple. Why?
Gemini Data’s Trademarks Precede Google’s By Years
Gemini Data has been developing artificial intelligence software since 2013, with the goal of helping users access AI-synthesized research through natural language searches. Basically, they want to help make AI research tools as easy to use as Google search, circumventing the need for computer engineer-level knowledge.
Not surprisingly, Google would also like an AI search tool that is as easy to use as its own search engine. Google began its endeavors under the Large Language Model LaMDA in 2020, according to this article tracing its AI history. However, the tech giant only began developing conversational AI in 2023, switching names from Atlas to Bard – then switching a year later to Gemini, which brings us to the current trademark conflict.
In their complaint against Google, Gemini Data claims that they registered trademarks for their business and AI tools “promptly.” In reality, their trademark registrations for GEMINI (6389900) and GEMINI DATA (6380947) were filed in 2019. The dates of first use listed for the different classes of goods are June 2017 for Class 9 and July 2019 for Class 42. Google did not file a trademark application for its Gemini AI tool until February 2024. Therefore, Gemini Data has a strong priority in the commerce argument.
How Did Google Miss This Possible Trademark Infringement? Maybe They Didn’t
Gemini Data asserts that Google must have conducted a trademark clearance search with its ample resources, so they conclude that the company must be wilfully infringing with a replica name. While mistakes happen, one does wonder how Google could have missed this.
Unsurprisingly, the USPTO refused Google’s trademark application and issued an Office Action. However, Google ignored this refusal, forging ahead with its rebrand of Bard as Gemini. It was this continued infringement that drove Gemini Data to file a lawsuit demanding a jury trial, delineating four counts:
- Federal Trademark Infringement under the Lanham Act
- False Designation of Origin and Unfair Competition
- Trademark Infringement and Unfair Competition under common law
- Unfair Competition under a California business and professional code
As we have covered before, trademarks of the same name can be registered by businesses in different classes – basically, trademark categories divided by industry. Because there are simply not enough words in any language for each business to have a wholly unique name.
However, in this situation, it is difficult to see how Google will justify its usage of “Gemini.” Its AI chatbot is not only in the same industry as Gemini Data – it is the same branch of generative AI within that industry.
Some in the world of trademarks might argue that – with so much overlap, and knowing Google’s sufficient legal sources and net worth in the trillions – it seems like the tech giant is trying to throw around its heft, hoping to outlast or settle with Gemini Data in a legal fight.
Others might recognize the possibility that the marketing team simply did not conduct a trademark clearance search and did not communicate well with the legal team; therefore, a big mess has developed. I have encountered this before, and it takes some time to work through the failed communications and resulting lawsuit.
If Google is trying to throw its weight around, Gemini Data’s team will need the stamina and finances to pursue the case to trial, although that rarely happens with trademark infringement cases – especially if one party has clear priority of use in commerce.