There have been several cases of voice AI agent developers losing money because their agents received automated calls. In web/public-facing you can implement rate limits, recaptchas, etc. but in real production environments? I am sharing a demo video of how you can integrate your voice AI agent with Hiya's AI Voice Detection solution that detects if a voice is real or not. This can be used to ping the AI agent with Vapi's Live Call control to let the agent know they are not talking to a human. Documentation on the solution: https://developer.hiya.com/docs/audio-intel/endpoints/authenticity-verifications-streaming Demo video: https://www.loom.com/share/7d44d31c70294354894821f995f1cc04?sid=a0775b96-ae12-4d06-aca5-a3abedce0e15 https://cdn.discordapp.com/attachments/1423334185115975813/1423334185300394129/image.png?ex=68dfeec2&is=68de9d42&hm=864a10d427730e2a5f7ebb62e1eb3318eb3364677d51c43e7dbf8bce26657727&