Innovative Technology for Eavesdropping Conversations

Researchers at Penn State University have developed a groundbreaking technology that uses millimeter-wave radar and artificial intelligence to eavesdrop on phone conversations from distances of up to 3 meters. The system can recognize and transcribe conversations with an accuracy of about 60%, analyzing the faint vibrations produced by a phone's receiver during calls. This is reported by Interesting Engineering.

The research is based on the adaptation of the open-source speech recognition model Whisper, which has been retrained on radar data to improve speech recognition quality even with noisy and low-quality data. The radar was set up approximately three meters from the phone, and the collected data was processed by AI capable of recognizing up to 10,000 words.

While the system's accuracy is not perfect, even partial recognition of key words can pose significant privacy risks. Researchers compare their technology to lip reading — it does not capture every word but helps understand the essence of conversations, especially when considering context.

This technology raises serious concerns about potential privacy threats, as malicious actors could exploit it for remote eavesdropping on personal conversations without the user's knowledge. The researchers emphasize that their work aims to raise public awareness about these risks and contribute to the development of protective measures against such forms of spying.

The study is supported by the National Science Foundation of the USA, and the team plans to continue its work on protecting personal information from emerging technological threats.