Ribbon Communications (Nasdaq: RBBN) has introduced Ribbon Call Trust Deepfake Detection, a network-based security solution designed to identify AI-generated synthetic audio during live telephone calls. Customer laboratory trials are expected to begin in Q1 2027.

The solution analyzes call streams through Ribbon Session Border Controllers (SBCs) using patent-pending algorithms. It is designed to detect synthetic speech, including short artificial segments inserted into otherwise legitimate calls, even after audio traverses multiple codecs or networks. No endpoint application is required, and detection information can be coordinated with third-party security systems.

Ribbon positions the offering as an extension of its Call Trust portfolio. Citing CrowdStrike, Ribbon says vishing intrusions doubled in the first half of 2026 versus the second half of 2025; it also cites Gartner research indicating 41% of surveyed CISOs encountered a deepfake social-engineering incident during an employee audio call in the preceding year. These figures have not been independently verified.

“AI has dramatically lowered the barrier to human impersonation,” said Sam Bucci, COO and EVP at Ribbon. Ribbon has not disclosed pricing, commercial availability, or independently validated detection accuracy.

Analysis

Network-based deepfake detection addresses a limitation of conventional caller authentication: verifying caller ID does not establish whether speech is genuine. STIR/SHAKEN authenticates caller ID information in supported IP networks, but does not assess whether the media stream contains AI-generated audio. SBC-based analysis could extend coverage across endpoints that cannot run detection software.

Performance under compression, transcoding, packet loss and background noise will be central to operational evaluation. Ribbon has not disclosed false-positive rates, detection latency or computational overhead. Planned Q1 2027 trials may help establish these metrics and the potential for managed voice-security services.