Judge dubious on taking trademark claims against OpenAI to trial

OAKLAND, Calif. (CN) — A federal judge on Tuesday indicated that OpenAI may have the stronger case when pitted against a similarly-named tech company on trademark infringement claims.

In the ongoing legal battle between OpenAI and the similarly named Open Artificial Intelligence, U.S. District Judge Yvonne Gonzales Rogers told attorneys that she’s leaning toward ruling for the ChatGPT creators on a recent motion that could decide the case before trial.

“At this time, I’m inclined to grant the plaintiff’s motion,” the Barack Obama appointee told attorneys at a hearing.

In a recent motion for summary judgment, OpenAI asked the judge to find in its favor on all counts of the almost two-year-old lawsuit, including five claims for trademark infringement and fraud against Open Artificial Intelligence and seven counterclaims that the ChatGPT creators are facing in response.

In the courtroom, the lesser-known tech company fought vigorously to get its case before a jury, calling the case “ripe with factual disputes.”

“Issues of credibility like this should go to the jury,” said Laura Chapman of Sheppard Mullin, who represented Open Artificial Intelligence.

Rogers said she wasn’t afraid to take the issue to trial, but that she wouldn’t do so without a proper amount of evidence.

“You have to have enough evidence to get to a jury. That’s why we’re here today,” Rogers told the defendant.

Open Artificial Intelligence’s founder, Guy Ravine, claims that he conceived the idea that became OpenAI in 2014 after talking to “top industry leaders,” and that OpenAI CEO Sam Altman and its founders stole those foundational principles to create the world’s most powerful AI company.

However, Ravine’s original idea for the company was supposedly much different, envisioned as a “Wikipedia-type platform for AI-related topics” where users could freely trade knowledge, post videos and start discussions.

Ravine turned down offers from OpenAI’s founders to buy the domain in both 2015 and 2022, also telling Altman that he would give up the website only if OpenAI donated millions of dollars to an “academic collaboration.”

However, Rogers pointed out that while Ravine’s ideas rang more like “a bunch of musing,” Altman and his co-founder, Greg Brockman, actually executed their ideas in a business setting.

“If you sit in your garage, and you’re making the first computer — until you make it commercial, you’re just sitting in your garage,” Rogers said.

Rogers said that this commercial use was an important factor, and that judging from the traffic on Ravine’s original website iteration of his idea, it didn’t seem like it had much traffic from users outside of himself and his employees.

“There is no evidence that anyone who ever shared a picture or made an edit as a bona fide user,” said attorney Margret Mary Caruso of Quinn Emanuel, representing OpenAI.

OpenAI also accused the company of falsifying documents submitted to the United States Patent and Trademark Office to receive its trademark. Specifically, they say the documents misrepresented that Ravine conducted commerce with his project as early as March 2015.

“Mr. Ravine’s declaration established that this website did not exist until 2016. So what are we talking about?” said attorney Robert Feldman, also of Quinn Emanuel, for OpenAI.

The smaller company countered that this was an honest mistake, and it should be up to a jury to decide if Ravine was just doing his best as a non-lawyer to navigate the “haze of instructions” on the trademark office’s website, or if he actually tried to mislead them.

“The standard for fraud is very high. It involves proving the applicant had a subjective intention to deceive the USPTO,” Chapman said**.**

Rogers added that she may hold an evidentiary hearing before proceeding to trial.

“You are asking me to use the power of this office to summon people from their daily lives and jobs to resolve a dispute,” Rogers told the defendants. “I’m not going to do that unless there are actually material disputes of fact to address.”

Rogers did not indicate when she would issue a ruling but said that she would stay away from addressing damages for now while she tries to determine liability.

Attorneys for both sides declined to comment on the proceedings.

OpenAI claims it’s trying to stop another artificial intelligence company from using its name and trademark, saying that the competitor has no proof that Altman stole the idea for his billion-dollar business.

OpenAI claims that after it announced its founding in 2015, Ravine snapped up a similar-sounding website and later changed his company’s branding and marketing to trade off its success, confusing consumers and harming its reputation.

“Defendants should not be allowed to profit from their misappropriation of OpenAI’s established name, leverage for themselves OpenAI’s goodwill, or fraudulently divert public interest in and demand for OpenAI’s products,” the ChatGPT creator said in its lawsuit.

Rogers, earlier this year, ordered a preliminary block on Ravine’s use of the trademark “Open AI” in commerce, saying that Open Artificial Intelligence’s branding was likely to confuse consumers.

In that order, she rejected Open Artificial Intelligence’s claim that its trademark rights trump and precede OpenAI’s and said that Ravine’s 2015 website appeared to be either inoperative or created just before applying for a trademark.

OpenAI primarily develops artificial intelligence models and applications, including the ChatGPT chatbot and an automatic speech recognition system. Both companies are Delaware-based corporations that mainly operate out of their headquarters in San Francisco.