OpenAI v. Open Artificial Intelligence Inc. is a case currently in the Northern District of California that pits an AI David versus an AI Goliath — but in a classic reverse trademark confusion case.
OpenAI is the owner of the popular ChatGPT service, and Open Artificial Intelligence, Inc. (OAI) is a company that sprung up way back when talk of AI was still aspirational, in 2015. This was a motion by OpenAI to dismiss the counterclaims of the lesser known defendant.
Defendant OAI asserted that it had common law rights, starting in 2015, before OpenAI launched ChatGPT in 2022. Then, suddenly in 2022, OAI was flooded with phone calls and emails, looking for assistance in using ChatGPT.
The defendant’s counterclaim is for reverse confusion. That’s where the “senior” user (who was first to commence its continuous use) is not as well known as the “junior” user, and seeks to enforce its common law (unregistered) trademark rights to stop the more popular party from using a confusingly similar trademark (here, virtually identical).
OpenAI argued that, in order for a senior user to argue reverse confusion without having had a registered trademark, OAI would have to demonstrate that, despite not being known well, they had acquired secondary meaning. The judge said OAI had sufficiently pled in their counterclaims that their secondary meaning had accrued among a very small group of people in 2015: the nascent group of individuals who were involved in artificial intelligence. In that community, the defendant argues, they had acquired secondary meaning. And that was enough to survive this motion to dismiss.
The other thing to consider, the judge said, was that instances of actual confusion were sufficiently alleged – in the form of receiving 1,100% more hits on their website, all asking for help regarding ChatGPT (which is the junior user’s product rather than the defendant’s). As a result, this documented actual confusion also supported OAI’s surviving the motion to dismiss its counterclaims.
The defendant also tried adding a counterclaim to hold the owners of OpenAI personally liable. To support this, they alleged that they had approached the owners, individually, and offered to sell OAI and its name, but they were turned down. The court said that’s all well and good, but that doesn’t mean that the owners of OpenAI engaged in intentional infringement of the trademark, or that they were the controlling, leading factor in engaging in this infringing conduct – a threshold that is required to pierce the corporate veil. Rather, a corporation’s structure is intended to protect individuals from liability like this, the judge said. As such, the counterclaims for individual liability were dismissed.
OpenAI might have avoided this lawsuit by making a business decision early on, instead of waiting and eventually suing. It’s potentially a big problem if the much smaller company prevails on demonstrating that they have enforceable trademark rights. That is, it’s conceivable that OAI shuts down the plaintiff’s use of its trade name OpenAI – which suggests that the relatively unknown senior user now has leverage for a better settlement.
