OpenAI Reveals Voice Engine: How A Trademark Filing Hinted At Product, Why There Could Be Future Risks

A common theme from 2023 to early 2024 has been the continued rise of interest in the use cases for artificial intelligence.

Among the leaders in the space is OpenAI, which unveiled a new product Friday.

What Happened: OpenAI took a large investment from technology giant Microsoft Corp MSFT in early 2024.

With the investment, Microsoft has solidified its plans to aggressively grow in the artificial intelligence space.

OpenAI is a leader in generative artificial intelligence and its ChatGPT chatbot can help turn text and prompts into stories, charts, tables, information and more.

On Friday, OpenAI unveiled its newest product, which is called "Voice Engine."

According to TechCrunch, Voice Engine is a voice cloning tool that expands on its existing text-to-speech API.

The company has been working on Voice Engine for around two years and it could allow users the ability to copy a voice after uploading a 15-second voice sample.

The announcement of Voice Engine by OpenAI on Friday follows a recent trademark filing shared by trademark attorney Josh Gerben of Gerben Law.

"OpenAI has filed a trademark application for: ‘Voice Engine,'" Gerben tweeted. "The filing claims that OpenAI will soon let users: 1. Build digital voice assistants. 2. Generate audio in response to user prompts."

The full trademark filing lists the following use cases for Voice Engine:

Automatic speech and voice recognition and generation

Building digital voice assistants

Generation of audio/and or voice in response to user prompts

Creating and generating voice and audio outputs based on natural language prompts, text, speech, visual prompts, images and/or video

Related Link: Microsoft Could Face Justice Department, FTC Probes Over Investment In ChatGPT Parent OpenAI

Why It's Important: TechCrunch is quick to point out the potential risks of the new voice cloning tool, which comes as deepfake technology has risen and AI could continue to provide risks of widespread fakes of celebrities and politicians confusing the public.

OpenAI is ensuring that the tool was created responsibly.

"We want to make sure that everyone feels good about how it's being deployed — that we understand the landscape of where this tech is dangerous and we have mitigations in place for that," OpenAI product staff member Jeff Harris told TechCrunch.

The current Voice Engine doesn't allow customization of tone or pitch for the voice. Prices seen by TechCrunch show a cost of around $1 per hour for the voice technology, which would be cheaper than voice actors.

Voice Engine is being released initially for a small group of developers and not publicly available at the time.

Read Next: Deepfake Dilemma Revived: Jordan Peele’s 5-Year-Old Obama Warning Gains New Relevance In 2024 Election Climate

Photo: Shutterstock