55/100

Moderate evidence

GainPublished recordCrimeGood Space

Neural Defend ranked #1 in the world for audio deepfake detection

Truvace record
TRV-2026-2876raw data (JSON)
Version
v1Published
First published
Evidence type
Benchmark or lab evaluation
Sources
1 sourceJournalism
sha256
b410b12b8fb7…check it

New Delhi [India], November 12: Neural Defend has been ranked number one in the world for detecting AI-generated voices. On an independent, publicly scored benchmark run by voice-AI evaluation firm Podonos, its Arcas 290M and Arcas 90M models hold first and second place among 28 detection systems, each scoring 99.93% on 4,524 audio clips. VMPL New Delhi [India], November 12: Neural Defend has been ranked number one in the world for detecting AI-generated voices. On an independent, publicly scored benchmark run…

The gain

On a Podonos benchmark of 4,524 audio clips, Neural Defend's Arcas 290M and 90M detectors each made three errors (99.93%), while Pindrop scored 95.05%, Hive 83.53% and Reality Defender 71.27%, according to a company press release.

What this doesn’t fix

The article is a syndicated advertorial press release from VMPL, and the results come from one benchmark of 4,524 clips rather than from live fraud-detection use.

Sources

1 on file
  1. The Tribunetribuneindia.com
    Journalism2026-11-12

Cite this record

v1

Truvace Impact Record TRV-2026-2876, v1: “Neural Defend ranked #1 in the world for audio deepfake detection.” Truvace, 2026-11-13. /record/TRV-2026-2876 (accessed on the day you copy it). sha256 b410b12b8fb7148a…

Change history

1 version

Every change made to this record since it was first published, and why. Nothing has changed since it went up.

  1. Publishedv1b410b12b8fb7…

    Certified into the record

Check this record hasn’t been altered

Check this record hasn’t been altered
How to check it yourself, without trusting this page

Download any version from /api/record/TRV-2026-2876 and hash its canonical field yourself, for example with shasum -a 256. You should get its content_hash. Each version’s text also ends with prev:and the hash of the version before it (version 1 points to 64 zeros). If anyone had changed a single character of any version after it was saved, those hashes wouldn’t line up.