Public methodology updates, casebook entries, and field notes from the team. Written by the people who built the detector.
A step-by-step methodology for verifying suspicious audio. Includes a fast checklist, the acoustic signs of synthesis, and how to validate using the detector.
Deepfakes are three problems, not one. How to choose a detector for voice, image, and video.
What "best" really means in voice detection. The 10 criteria to compare detectors by.
From CEO ransom scams to election interference. Real categories of harm AI voices already enable.
Plain-English explainer of what the detector is, why it exists, and how it works.
The 2024 election cycle was the first where AI-generated robocalls reached voters at scale.
Executive-impersonation fraud in the CEO's real voice, and the controls that stop it.
A cloned voice makes an old con new. The red flags, and how to protect older relatives.
The most common voice-scam patterns. What attackers say, what they ask for, how to verify.
What still gives synthetic voices away on close listen, even with the latest generators.
The fundamentals: what an AI detector is, what it measures, and our approach.
A field note on a category of use we did not predict. Synthesized voices in religious settings.