Sierra launches fleming-1 to detect when an AI agent is calling
Sierra released fleming-1, a model that analyzes a caller's speech in real time and scores the audio for signs it was generated by AI, flagging calls likely made by an agent. Sierra says the model is tuned to be conservative by default so real people are not inadvertently flagged, and that it provides information rather than judgment: businesses decide whether to add verification steps or simply measure how often agent calls occur. The model works with any voice agent built on Sierra and can be switched on. Sierra also said it announced Personal Agent Protocol, an open standard for personal agents to work with businesses in an authorized way, and noted its own agents began calling each other in healthcare last year.
As personal agents spread, companies need to know whether a caller is human or AI, and this kind of detection could become the caller ID of the agent era.