OpenAI files trademark application for GPT-5

The trademark application suggests OpenAI may be working on a conversational voice assistant capable of translating live speech and generating its own AI models.

The San Francisco-based company behind ChatGPT has filed a trademark application with the US Patent and Trademark Office (USPTO) for GPT-5, according to the agency’s online registers website.

The documents show the application was filed on July 18th and have been accepted by the USPTO for examination.

The status indication dated August 2nd notes that the application “has met the minimum filing requirements” but is yet to be assigned an examiner. The application also shows no “current use” of the trademark but indicates “intent” to use it.

The most advanced version of ChatGPT released so far runs on the GPT-4 model and is available to paying subscribers.

According to the application, GPT-5 trademark registration is intended to cover “downloadable computer programs and downloadable computer software for using language models.”

Notably, the patent covers software for the “artificial production of human speech and text,” in addition to “natural language processing, generation, understanding, and analysis,” suggesting OpenAI’s plans for an AI-powered voice assistant.

It is further supported by the stated intent of the patent to cover the “translation of text or speech from one language to another” and “downloadable computer software for voice and speech recognition.”

The current version of the ChatGPT app supports voice recognition in multiple languages, meaning a prompt can be dictated vocally, but the chatbot does not have a voice of its own, Alexa-style. Currently, it only operates on text input and output.

The trademark application also suggests that GPT’s next iteration could generate other large language models, with mentioned services including “software for sharing datasets for the purpose of machine learning” and “software for developing and implementing artificial neural networks.”