ChatGPT maker OpenAI on Friday shared a preview of a new artificial intelligence (AI) tool that can generate “natural-looking speech” and mimic the human voice.
The tool, called “Voice Engine,” requires only “a single 15-second audio sample to produce a natural-sounding voice that closely resembles the original speaker,” OpenAI said in a blog post. Ta.
According to examples shared by the AI startup, speech engines can provide reading assistance, content translation, and audio to people who are nonverbal or suffer from language disorders. However, OpenAI acknowledged that the tool could pose “serious risks of particular concern during an election year.”
The company first developed its voice engine in late 2022 and began testing it privately with a “small group of trusted partners” late last year.
OpenAI emphasized that these partners agree to usage policies that require the original speaker's explicit and informed consent and prohibit impersonation of individuals without their consent.
The company noted that partners must also disclose that the audio was generated by AI and that audio generated by Voice Engine has watermarking capabilities to help track its origin.
The AI startup says widespread adoption of such tools will include voice authentication to “verify that the original speaker intentionally added the voice to the service,” and the creation of fraudulent voices. He stated that he believes it is necessary to include a “list of prohibited voices'' to prevent such incidents. A voice similar to a celebrity.
OpenAI also recommended that institutions phase out the use of voice-based authentication to access bank accounts and other sensitive information.
The company still seems somewhat uncertain as to whether it will eventually release this tool more broadly.
“We want to start a conversation about the responsible adoption of synthetic speech and how society can adapt to these new capabilities,” OpenAI said in a blog post. “Based on the results of these conversations and small-scale testing, we will make more informed decisions about whether and how to deploy this technology at scale.”
The new audio technology comes amid growing concerns about the potential for AI-generated deepfakes to spread election-related misinformation.
Earlier this year, a message imitating President Biden was sent to New Hampshire voters ahead of the January primary, urging them not to go to the polls.
Steve Cramer, a veteran Democratic operative, later admitted to creating the fake robocalls and said he did so to draw attention to the dangers of AI in politics.
A local Arizona newsletter similarly published an AI-generated deepfake video of Republican Senate candidate Kali Lake last month to alert readers to “how good this technology is.”
Copyright 2024 Nexstar Media Inc. All rights reserved. This material may not be published, broadcast, rewritten, or redistributed.