Megafon’s virtual assistant Eva has upgraded its deepfake detection skills beyond phone calls – it now identifies synthesized speech in voice messages and video clips across messaging apps. The bot is free to use for anyone, regardless of their mobile carrier.
Users can forward suspicious voice content directly to Eva, which employs a multifactor model analyzing sound wave parameters to spot potential threats. One common scam involves an unknown contact posing as a friend and urgently asking for money transfers.
Previously, Eva could detect fake voices during live phone calls and automatically disconnect the line. Importantly, it differentiates scam bots from benign automated messages, like delivery notifications, avoiding unnecessary call interruptions.
Megafon’s commercial director Dmitry Rudskikh explained the expanded feature:
”Scams are becoming more sophisticated and personalized. Even the most robust technical defenses can’t fully eliminate social engineering risks. This is especially true for newer messaging apps, where users rapidly grow their contact lists and can’t always be sure a message is really from someone they know.”
Dmitry Rudskikh, Megafon
With digital scams growing increasingly clever, Eva’s ability to analyze voice clips in popular messaging apps adds a vital line of defense. Unlike Apple and Google, which focus more on content moderation and reporting tools in apps like iMessage and Google Messages, Megafon is integrating AI-driven voice authentication to spot fakes in real time. Eva’s free availability regardless of carrier aims to protect a broad user base in Russia’s fragmented telecom scene.
Experts and users should watch how Eva’s deepfake detection evolves amid rising voice-based scams. Its success may encourage other carriers and app makers worldwide to embed voice authentication features, raising the bar for scam prevention in messaging platforms.