Introducing ElevenLabs AI Speech Classifier

At its best, generative AI can boost human productivity, creativity and wellbeing. At ElevenLabs, we believe in the transformative potential of these technologies but we also recognize that to fully harness their benefits, we must ensure their safe and responsible use. We're now launching the Classifier to help prevent the spread of misinformation by making the source of audio content easier to assess.

Transparency over AI-generated content

The Classifier allows anyone to upload a sample and analyze whether it contains AI audio from ElevenLabs. It's the first proprietary mechanism of its kind in the audio space and it is a critical step in our mission to develop efficient tracking for AI-generated media. With today’s launch, we seek to further reinforce our commitment to transparency around

Mechanism and known limitations

Audio generated by our models has certain detectable characteristics. When you upload an audio sample to the AI Speech Classifier, our algorithm will scan for them to assess whether the content was indeed generated by our platform, currently maintaining >99% accuracy if the input was unmodified. If it underwent Codec or reverb transformations, the Classifier is over 90% accurate. This figure drops the more the content has been post-processed. Adding more audio tracks will also affect the result.

We are continuously analyzing and improving our model to be able to detect more audio transformations and we expect these outcomes to improve.

Test the Classifier yourself:

Call for input

We are releasing the Classifier publicly today to foster AI detection initiatives. We want to enhance transparency over AI-generated content by starting with our own. A plethora of AI content from other commercial and open-source models remains undetected. We encourage others to join our effort, and we encourage reporting suspected cases as we continue to improve the Classifier’s accuracy and detection capabilities to cover more audio transformations, as well as content generated on other platforms, including both commercial and open-source models. We are eager to collaborate with interested partners on deeper API integration and enhancing the value of this tool.

Please connect with us if you're interested in either partnership or integration.

A commitment to transparency

Our goal at ElevenLabs is to produce safe tools that can create remarkable content. We believe that our status as an organization gives us the ability to build and enforce safeguards which are often lacking in open source models. With the AI Speech Classifier launch we also aim to empower businesses and institutions to leverage our research and technology to bolster their respective safeguards.

Read our safety page to learn more about AI safeguards at ElevenLabs.