Truecaller, the Swedish caller ID service, has upgraded its AI Assistant to answer calls using a digital version of the user's voice. This enhancement is made possible through a partnership with Microsoft, leveraging the Personal Voice technology from Azure AI Speech.
Capabilities of the AI Assistant
The AI Assistant, introduced in 2022, already offers features such as call screening, message reception, automated responses, and call recording. The new feature will allow users to create a digital replica of their voice by recording a short script during the setup process. This digital voice will then be used to interact with callers.
Microsoft's Personal Voice technology, showcased at its Build 2024 developer conference, plays a crucial role in this feature. Truecaller, as an early access partner, plans to roll out this capability in the coming weeks across several countries, including the US, Canada, India, Australia, South Africa, Sweden, and Chile. The AI Assistant will notify callers that they are speaking with a digital voice.
Security and Verification
To address security concerns, Microsoft has implemented watermarks in the speech outputs of Azure AI Speech's personal voice feature. These watermarks ensure that the synthetic voices can be verified by identification tools. Additionally, this feature is available only through registration for developers interested in creating applications with it.
Market Rollout and Availability
Truecaller's new feature will be available to paid users initially. Users will need to record a short script to create their digital voice, following their consent. While the introductory greeting template cannot be modified, follow-up responses can be customized to suit individual preferences.
The rollout will begin with public beta users before expanding to all eligible markets. Raphael Mimoun, Truecaller's product director and general manager, expressed confidence that this feature will significantly enhance call management for users.
