Vastav AI offers India-built deepfake checks for images, video, and audio

TV9 Hindi reported on Vastav, a deepfake detection system developed by Zero Defend Security. The system is described as checking AI-generated photos, videos, and audio, then producing reports that use confidence scores, heatmaps, and metadata insights to show why content may be fake. The report says Vastav was recognized at CID Karnataka's CIDECODE hackathon and claims up to 99% accuracy. That accuracy claim should be treated as vendor-reported until independently benchmarked, but the product is still worth tracking because India's deepfake harm stories increasingly need practical forensic tools for police, media, and government users.

Key details: Published March 13, 2025 at 10:37 IST, Vastav was developed by Zero Defend Security, The system checks images, videos, and audio for signs of AI generation or manipulation, TV9 Hindi says it uses confidence scores, heatmaps, metadata insights, and detailed reports, The reported 99% accuracy is vendor-framed and needs independent benchmarking.

Why it matters: As deepfake cases move from celebrity harm to police work and evidence handling, detection tools matter, but their claimed accuracy needs independent testing before becoming trusted infrastructure.