xSentinel

With technological advances making it nearly impossible for humans to tell whether a voice is real or digitally generated, organizations need systems of their own to provide a line of defense against bad actors attempting to defraud their business. xSentinel, part of our AI.X family of deepfake defense solutions, uses patented technology to indicate if a voice is likely synthetic, minimizing fraud.

Early fraud detection
xSentinel sits at the point of entry for calls, providing an early and accurate signal as to whether the audio has been synthetically generated.

Futureproof with AI
The xSentinel algorithm keeps pace with the changing fraud landscape through machine learning, continually adjusting to potential new threats.

Plug and protect
A standalone service that integrates into any voice communication platform, xSentinel drops into existing workflows with little to no extra hardware/software investment.

Real-time analysis
xSentinel analyzes incoming calls as they stream in real time, giving results to your agents/IVR within seconds so any necessary steps can be taken quickly and confidently.

No added friction
With no enrollment or guided action required, xSentinel can go to work protecting your business immediately without inconveniencing your customers.

Universally functional
xSentinel analyzes technical aspects of the audio, not the voice or what is being said, eliminating barriers created by language or dialect, making it a globally effective tool.

Deepfake voice generators have rendered the ability of a human to discern if a voice they are hearing is real obsolete. Bad actors can now mimic the accent, age, gender, language, speech patterns, and even the actual voice of their target, making it impossible for an agent to know if they are talking to a person or a machine.

Compounding the issue, the combination of generative AI, deepfake, and voicebots creates the perfect storm of fraud threats to call and contact centers. With these tools, fraudsters have the ability to make thousands of synthetic calls per day that not only sound human but have all of the right knowledge-based authentication responses.

xSentinel gives the power of discernment back to businesses, with no need to analyze voice, language, or any other identifying feature of the call. Seconds after a caller begins speaking, multiple patented and proprietary algorithms detect various cues that indicate digitally generated audio. Within seconds, xSentinel provides a status and confidence score that, combined with the risk level of the call, can help determine whether the call should continue, further authentication should be initiated, or the call should be terminated.

xSentinel eliminates the advantages that modern voice generation technology can provide for bad actors by providing a signal to potential fraud quickly and accurately.

Get Early Access

Does xSentinel use voice biometrics?
No. xSentinel strictly analyzes the audio it processes to determine if it is a genuine person speaking or a digitally generated voice. It does not have the capability to identify the actual person. Our xVoice application provides voice biometric authentication.

Why do I need xSentinel if I already have biometric voice authentication?
xSentinel is the perfect complement to any authentication process, as it provides a signal that extra scrutiny may be required to confirm the authenticity of a caller. While highly unlikely, especially with replay protection in place, a well-made deepfake has the potential to fool biometric security. A fraudster who has gone to the effort of creating a deepfake, is likely to have also gathered answers to the common “what you know” authentication questions, potentially defeating a second authentication factor. With xSentinel, even an authenticated biometric with correct PII answers can be called into question and further steps can be taken to ensure the identity.

Does xSentinel require a Daon platform?
While xSentinel can run on IdentityX or TrustX, neither is required for it to be used by your organization. xSentinel can be connected to any voice communication system to analyze incoming calls in real time.