Deepfake Police

Check suspicious digital media content and detect deepfakes quickly and reliably.

Audio

Detect cloned voices and synthetically generated music.

Image

Verify image authenticity and stop misinformation before it spreads.

Video

Safeguard video conferencing and strengthen identity verification.

Submit a Link

Paste a link to any publicly accessible media file. Supportsand more.

Upload a File

Upload audio, image, or video files directly from your device.

Try Now

Motivation

The Fading Line Between Reality and AI

Choose the real kitty.

How good are your detection skills?

Start the Challenge

About

How It Started

This project emerged from university research in advanced Deepfake detection. In recent years, research has made significant progress, and highly capable detection models already exist. Yet these technologies rarely reach the people and organizations who need them in everyday digital life. We are closing that gap.

Our mission is simple:

Deepfake detection for everyone.

We transform our research and detection models into an accessible platform that delivers practical analysis for real-world users.
But detection is only one part of the solution.
We support people in navigating a media landscape where synthetic content is becoming increasingly realistic and widespread. That means not only building detection technology, but also raising awareness of emerging deepfake trends, strengthening people’s ability to recognize manipulated media through education and hands-on challenges, and enabling professional use of our detectors through APIs.

Roadmap

The Journey Ahead

What we are currently working on

First release of our image deepfake detection

Preparing our service to reliably analyse images for signs of manipulation and and make this feature available to users for the first time.

Service stability & reliability

Ensuring the service runs smoothly, remains available, and delivers consistent results, even as usage grows.

What we plan to do next

Audio deepfake detection

Expanding our technology to also analyse voices and audio clips, helping identify AI-generated or manipulated speech.

Social media detection bots

Automated bots that monitor trending posts and respond when tagged. They can analyse suspicious content and reply with detection results to help inform the public.

Professional API access

A secure interface that allows organizations and other software platforms to connect directly to our deepfake detection service and use it within their own tools and workflows.

FAQ

Questions Answered

What is a Deepfake?

Deepfakes are synthetic audio, images, or videos that appear realistic but have been generated or altered by deep neural networks, commonly referred to as generative AI models. While these models have legitimate applications, their rapid improvement, growing popularity, and widespread availability also create opportunities for misuse.

One form of deepfake involves imitating individuals without their knowledge or consent, which can enable various criminal activities.
Deepfakes are also used to target large audiences with misinformation, contributing to an erosion of trust in digital content overall.

What kinds of Deepfakes can be reliably detected?

Our detectors identify only content that has been artificially generated or manipulated. They support audio, image, and video inputs, provided that a significant portion of the media has been fabricated by a generative AI model.

Detection reliability depends strongly on input quality:

Higher-resolution, minimally compressed files typically yield higher confidence and more accurate predictions.Low-resolution, heavily compressed, or very short inputs will most likely result in low confidence, and in such cases the prediction is more likely to be incorrect.

How do we detect Deepfakes?

We examine digital media content to identify fingerprints, meaning subtle and recurring patterns that appear when content is generated or altered by AI.

A helpful comparison is traditional forensic work:

Just as a camera’s optical lens or sensor chip leaves tiny, characteristic traces in every photo it captures, generative models leave identifiable patterns in the output they produce.

Our detectors can analyse content far more thoroughly than a human could ever do. They extract and evaluate a large number of carefully engineered features across different representations and scales, and combine them to reach a prediction. This allows us to uncover inconsistencies and patterns that are invisible to the human eye or ear.