OpenAI moves to trademark its ‘reasoning’ models

OpenAI has filed a trademark application for its latest AI model, o1, as the firm moves to shield its IP.

On Tuesday, OpenAI submitted paperwork to the U.S. Patent and Trademark Office (USPTO) to register the trademark “OpenAI o1.” Interestingly, the documents reveal that OpenAI filed for a foreign trademark application in Jamaica in May, months before o1 was announced.

The USPTO hasn’t granted OpenAI the trademark yet. According to the office’s online database, the application is currently awaiting assignment to an examining attorney.

OpenAI has said that it intends for o1, its first “reasoning” model, to expand into a series of models trained to perform complex tasks. Unlike most models, reasoning models effectively fact-check themselves by spending more time considering a question or query — helping them avoid some common AI pitfalls.

OpenAI, which has filed for around 30 trademark registrations to date, including for “ChatGPT,” “Sora,” “GPT-4o,” and “DALL-E,” famously failed to trademark “GPT” in February after the USPTO ruled that the term was too generic. GPT, which stands for “Generative Pre-trained Transformer,” was in use in other contexts and by other companies when OpenAI submitted its request, the office noted.

OpenAI hasn’t aggressively asserted its trademarks yet — save for one. For several months, the startup has been fighting technologist and entrepreneur Guy Ravine for the right to use “Open AI,” which Ravine claims he pitched as a part of an “open source” AI vision around 2015 — OpenAI’s founding year.

A federal circuit court upheld a preliminary injunction in OpenAI’s favor early this fall, ruling that OpenAI was likely to prevail against Ravine.

Subscribe for the industry’s biggest tech news