ElevenLabs has introduced the AI Speech Classifier, a tool designed to enhance transparency and safety concerning AI-generated audio content by allowing users to upload and analyze samples to determine if they contain ElevenLabs' AI audio. This proprietary mechanism is the first of its kind in the audio space, aiming to prevent misinformation by making audio sources easier to assess. The Classifier boasts over 99% accuracy for unmodified inputs and over 90% accuracy for those with codec or reverb transformations, though accuracy decreases with more post-processing. ElevenLabs is committed to improving the system to detect additional audio transformations and encourages collaboration with partners to expand its capabilities to cover content generated by other platforms. The initiative reflects ElevenLabs' dedication to developing safe and innovative tools, enhancing transparency, and empowering businesses and institutions to strengthen their safeguards against the misuse of AI-generated content.