ElevenLabs uses an AI-based content moderation system that embeds inaudible watermarks in all audio it generates and offers a public AI Speech Classifier tool that detects whether an audio clip was created using ElevenLabs' technology.

ElevenLabs embeds inaudible watermarks into audio files at the frequency level during the generation process. The AI Speech Classifier tool, accessible via the ElevenLabs dashboard, accepts an audio file as input and returns a probability score indicating how likely the clip was generated by ElevenLabs. The company also operates platform-level moderation using AI classifiers and human reviewers to monitor for violations of its Prohibited Usage Policy. For Conversational AI agents, a live message-level moderation feature can be enabled to automatically drop calls when prohibited content is detected. ElevenLabs additionally partnered with Reality Defender in July 2024 to develop audio deepfake detection models.