The company said it is not yet widely releasing the technology, which could be subject to misuse.
04/09/2024 1:50 P.M.
3 minute read
A new tool from ChatGPT-maker OpenAI can create synthetic voices based on 15 seconds of recorded audio, the company shared recently.
OpenAI released samples from early tests of its new text-to-speech tool, which is called Voice Engine, as well as background on how the company developed it. However, OpenAI said it is not yet widely releasing the tool due to potential risks.
“We hope to start a dialogue on the responsible deployment of synthetic voices, and how society can adapt to these new capabilities,” OpenAI said in a blog post. “Based on these conversations and the results of these small-scale tests, we will make a more informed decision about whether and how to deploy this technology at scale.”
OpenAI is not the first company to develop an AI-generated voice service.
“Voice cloning tech in general is not particularly new—there have been several AI voice synthesis models since 2022, and the tech is active in the open source community with packages like OpenVoice and XTTSv2,” according to an article in Wired. “But the idea that OpenAI is inching toward letting anyone use its particular brand of voice tech is notable.”
Voice-cloning technology can be used for a variety of criminal purposes, such as to break into bank accounts that use voice authentication or to trick consumers into giving out sensitive information.
OpenAI cited certain actions it hopes to see from policymakers, researchers and developers to prevent synthetic voice misuse:
- “Phasing out voice-based authentication as a security measure for accessing bank accounts and other sensitive information
- Exploring policies to protect the use of individuals’ voices in AI
- Educating the public in understanding the capabilities and limitations of AI technologies, including the possibility of deceptive AI content
- Accelerating the development and adoption of techniques for tracking the origin of audiovisual content, so it’s always clear when you’re interacting with a real person or with an AI.”
ACA’s Take
The Federal Communications Commission is focused on “voice cloning scams” to stop bad actors, ACA International previously reported. In February, it adopted a declaratory ruling confirming that the Telephone Consumer Protection Act’s restrictions on the use of “artificial or prerecorded voice” include artificial intelligence technologies that mimic human voices and/or create calls with a prerecorded voice.
“As a result, callers that use such technologies must obtain the prior express consent of the called party to initiate such calls absent an emergency purpose or exemption,” the FCC states in the ruling (PDF), which is now in effect.
ACA members using AI technologies that replicate human voices on outbound calls will need prior express consent of the called party, similar to the TCPA’s requirements on using any automatic telephone dialing system or prerecorded voice call. If you already have consent to make prerecorded or artificial voice calls, no additional consent should be required.
The FCC’s ruling also clarifies all AI or prerecorded voice messages must provide certain identification and disclosure information for the entity responsible for initiating the outbound call, consistent with current obligations around prerecorded or artificial voice calls.
Additionally, telemarketing or advertisement calls using the technology need to include opt-out methods. These requirements are applicable to any AI technology that initiates any outbound telephone call using an artificial or prerecorded voice to consumers.
You can listen to ACA’s CEO Scott Purcell and Vice President of State Government Affairs Andrew Madden discuss the FCC’s AI ruling on a recent episode of the Credit and Collection News Podcast.
The ACA Huddle on April 10 also covers an overview of consumer consent on text messages, ringless voicemails, and emails. David Kaminski, managing partner at Carlson & Messer LLP, will cover the legal frameworks, best practices, and emerging trends surrounding consent. The ACA Huddle is at 11 a.m. Central. You can register and find recordings archives on the ACA Huddle website.
Remember, subscribe to ACA Daily and Member Alerts under your My ACA profile when logged in to acainternational.org to receive updates on the ACA Huddle.