ChatGPT: How do I win a trademark lawsuit?
It wasn’t long after ChatGPT’s public launch in November 2022[1] that headlines hailed it as a revolutionary product,[2] and it set a record for the fastest-growing consumer application in history.[3] Since that time, generative artificial intelligence has upset industries[4], changed how people write and research[5], and it’s made OpenAI, ChatGPT’s parent company, worth an estimated $260 billion.[6] Despite its success, however, OpenAI finds itself embroiled in a fight for its very name as a lesser known company called Open Artificial Intelligence claims OpenAI is infringing on its name.[7]
It’s a cross-town battle between two San Francisco based tech companies, and one that, if OpenAI finds itself on the losing side of, could mean a costly hoisting of its own petard as it was OpenAI that initially brought the suit.[8] But now, the company is fending off a series of counterclaims that have survived OpenAI’s efforts to dismiss.[9] So, here’s a 1ook at OpenAI’s trademark lawsuit:
OpenAI’s Complaint
On Aug. 4, 2023, OpenAI filed a complaint in the Northern District of California against Open Artificial Intelligence alleging federal and common law trademark infringement, among other claims.[10] The suit came after OpenAI was denied registration of a standard character mark for its name “OpenAI,”[11] allegedly thanks to a letter of protest filed with the United States Patent and Trademark Office by Guy Ravine, the president of Open Artificial Intelligence.[12] The letter notified the USPTO of a trademark registered in August 2017 for the standard character mark “Open AI.”[13] After a review, the USPTO denied OpenAI’s application, finding that “[t]he applicant’s mark is essentially identical to the registrant’s mark” and that the addition of a space between the words “open” and “AI” does not change the “commercial impression of otherwise confusingly similar marks.”[14] It was following this decision, that OpenAI then sued Open Artificial Intelligence, alleging that Open Artificial Intelligence does not possess a valid trademark, and even if it does, that OpenAI’s mark is senior to it, and as such, Open Artificial Intelligence has infringed on OpenAI’s intellectual property.[15]
In alleging that Open Artificial Intelligence does not have a valid mark, OpenAI alleges that Ravine filed for his trademark the same day that OpenAI co-founders Sam Altman and Greg Brockman announced their company, Dec. 11, 2025,[16] and that Ravine supplied “fraudulent specimens” in order to make it appear as though Ravine had used the “Open AI” name in commerce prior to filing for the trademark.[17] But even if Ravine’s trademark is valid, OpenAI nevertheless claims that Open Artificial Intelligence’s “selection, adoption, and use” of its Open AI mark is intended to trade on OpenAI’s name and reputation.[18] OpenAI alleges that for years, Open Artificial Intelligence’s domain, open.ai, redirected to OpenAI’s website as late as January 2022.[19] Additionally, OpenAI alleges that Open Artificial Intelligence created similar subdomains to those used by OpenAI, which OpenAI claims is evidence of Open Artificial Intelligence’s bad faith efforts to profit off OpenAI’s marks.[20] For example, OpenAI alleges that it began using the subdomain chat.openai.com for its ChatGPT chatbot, and that roughly four months later Open Artificial Intelligence created its own subdomain under chat.open.ai.[21] OpenAI is seeking permanent injunction against Open Artificial Intelligence, disgorgement of any profits the company may have made as a result of its alleged infringement, and punitive damages.[22]
Open Artificial Intelligence Claps Back
OpenAI’s complaint is, of course, one side of the story as defendant Open Artificial Intelligence has retaliated with counterclaims of its own.[23] In rambling filings, which Judge Yvonne Gonzalez Rogers said spill “much proverbial ink pontificating about their view on artificial intelligence, most of which is not at all relevant to the actual legal claims,”[24] Open Artificial Intelligence claims to have started Open AI as an open-sourced non-profit AI organization “for the benefit of humanity” in San Francisco in 2014,[25] prior to the incorporation of OpenAI on Dec. 8, 2015.[26] Open Artificial Intelligence paints Ravine as someone who foresaw the power of “artificial general intelligence” (AGI),[27] and that his Open AI was an “ideological bombshell”,[28] which was “the secret sauce to build AGI collaboratively and in a manner that benefited humanity, not individual companies.”[29] Instead of Ravine’s Open AI revolutionizing AGI for the benefit of humanity, however, Open Artificial Intelligence claims OpenAI co-founders Sam Altman and Greg Brockman learned of Ravine’s Open AI and rushed to “usurp” it,[30] by duplicating Ravine’s “initiative, principles, mission, vision, words, domain, and even the Open AI name.”[31] Open Artificial Intelligence states that in response to the public announcement of OpenAI on Dec. 11, 2015, Ravine immediately filed an application to register his Open AI Mark with the USPTO in order to protect his intellectual property.[32] Ravine’s trademark application was eventually registered on Aug. 1, 2017.[33]
Clearing OpenAI’s Motion to Dismiss
The most significant hurdle in the litigation so far for Open Artificial Intelligence came in convincing Judge Rogers to allow its counterclaims to survive OpenAI’s motion to dismiss.[34] After initially taking issue with most of Open Artificial Intelligence’s claims,[35] Judge Rogers eventually ruled that Open Artificial Intelligence’s effort to sue Sam Altman and Greg Brockman in their personal capacities would fail,[36] but the infringement claims against OpenAI could move forward. In allowing the infringement claims to continue, Judge Rogers found that Open Artificial Intelligence had “provided substantive allegations of confusion” between Ravine’s Open AI and OpenAI, and said “the claim may proceed.”[37] Judge Rogers did note, however, “[w]hether defendants can actually garner evidence of [confusion] is a separate issue.”[38]
Where things stand
Following Judge Roger’s decision, the case has continued to move forward with discovery.[39] Additionally, OpenAI has now answered Open Artificial Intelligence’s counterclaims, asserting a number of affirmative defenses.[40] Among them are laches, as OpenAI alleges it has used its name since 2015, but neither Ravine nor Open Artificial Intelligence ever pursued enforcement to protect its mark until after OpenAI initiated the current lawsuit.[41] OpenAI also states that in multiple communications between Ravine and both Altman and Brockman dating back to 2015, Ravine and Open Artificial Intelligence never suggested OpenAI should stop using its name, which OpenAI claims supports a defense of waiver, estoppel, and acquiescence.[42] Among its other defenses, OpenAI also continues to dispute that Open Artificial Intelligence has a valid mark, alleging fraud through “attempts to manufacture trademark rights before the United States Patent and Trademark Office and this Court.”[43] Open Artificial Intelligence has not yet responded to OpenAI’s answer.[44]
Footnotes
