Hiya: Security – Real-Time Scam Protection with AI  - GSMA Foundry
Tuesday February 18, 2025

Hiya: Security – Real-Time Scam Protection with AI 

Resource thumbnail

Challenge: Generative AI has revolutionised voice-related scams, making fraudulent calls more sophisticated and harder to detect. Deepfake voice impersonation is on the rise, posing an increasing threat to businesses and consumers. Traditional fraud prevention measures are no longer sufficient, as scammers rapidly adapt their tactics to bypass existing defenses. Carriers need real-time, AI-driven scam protection to safeguard users from these evolving threats. 

Solution: Hiya’s Gen AI-powered scam protection identifies both human and AI-generated voice scams in real-time during phone calls and video conferences. The solution uses advanced voice verification technology to determine whether the voice on the line is authentic while simultaneously analyzing conversation patterns for scam indicators and language. 

If a potential scam or AI-generated voice is detected, the Hiya AI Digital Assistant issues an immediate warning to the user, providing the option to end the call or conference. By leveraging deep learning models trained on vast datasets of scam calls, Hiya ensures that even the most sophisticated AI-generated fraud attempts are flagged within two seconds with 99% best-in-class accuracy

Impact: Detects scams and AI voices in less than 2s. Generative AI-powered scam protection operates during a call to detect scams and AI voices in real time and provides a warning during the call to suggest you hang up if you have concerns. It can work in conjunction with Adaptive AI which would block a fraud call before it happens when detected based on it’s pre-call reputation. 

  • Proactive Scam Prevention: Real-time detection prevents voice fraud before financial or data losses occur. Generative AI-powered scam protection performance data is about speed and accuracy during the call,  
  • Consumer & Enterprise Protection: Safeguards both individuals and businesses from deepfake impersonation scams. Benchmarked at achieving over 99% accuracy in both controlled and real-world scenarios 
  • Industry-Leading Accuracy: AI-driven verification delivers a 99% detection rate with minimal false positives. 

Topics: Security, and customer experience 

Cost: Medium – Requires network integration but delivers long-term fraud mitigation benefits and revenue benefits. 

Location: Europe, North America, Latin America, Asia Pacific 

MNO: Hiya