Digital Forensics
Deepfake Detection Tools
Hive Moderation provides multiple deepfake detection tools capable of identifying deepfake images, audio, and videos.
Tag(s): Audio, Image, Video
DeepFake-O-Meter integrates a variety of deepfake detection algorithms capable of identifying deepfake images, audio, and video.
Tag(s): Audio, Image, Video
Image Whisperer combines multiple detection systems and LLM judgment to deliver a color-coded verdict with a clear explanation.
Tag(s): Image
Decopy AI is an AI image detection tool trained on images generated by Midjourney, Stable Diffusion, DALL·E and Flux.
Tag(s): Image
Resemble AI's audio deepfake detection tool allows user to upload, or provide a link to, an audio clip for analysis.
Tag(s): Audio
GPTZero is a detection tool focused on identifying text generated by a variety of Large Language Models (LLMs), like ChatGPT.
Tag(s): Text
Guides for Identifying AI-generated Content
Reporter's Guide to Detecting AI-Generated Content teaches how to identify AI-generated content and outlines seven categories of deepfake detection skills.
Synthetic Photography Detection provides evidence-based strategies for identifying AI-generated images of people, objects, indoor scenes, and outdoor scenes.
The Human Guide to Detecting AI Imagery supports users in detecting distinct visual artifacts that can help identify AI-generated content they may encounter online.
Content Credentials, Watermarking, and Reverse Image Search
Content Credentials are an open technical standard designed to track digital content's provenance.
Tip: Use original full-res image for best results.
Google’s SynthID embeds invisible watermarks directly into AI-generated images, audio, text or video. To use it, upload a file to Gemini and ask: “Is this image generated by AI?”
Google Lens allows users to perform visual searches to identify similar content across the web to establish provenance and legitimacy. Available as an app or through Google products.
Methods for Evaluating Information
The SIFT Method, developed by Mike Caulfield, can be used to help determine whether online content can be trusted as a credible or reliable source of information.
The C.R.A.P. Test, developed by Molly Beestrum, outlines key criteria and considerations users should apply when evaluating information online and accessing its credibility.
The RADAR Approach, developed by Jane Mandalios, provides a framework for evaluating the legitimacy of different information sources online.
Disclaimer: This page includes commercially available and experimental tools designed to detect synthetic or manipulated media (“deepfakes”). These tools operate with varying methodologies and levels of maturity. Performance may be limited, inconsistent, or subject to change as providers update algorithms, modify service tiers, introduce subscription models, or discontinue tools. Outputs should be considered tentative indicators rather than conclusive findings. Users are encouraged to corroborate results through independent analysis and to exercise caution when applying these tools in sensitive, legal, or high-impact contexts.