The InVID Verification Plugin is a browser-based verification toolkit designed to help users investigate and verify images and videos circulating on social media. It brings together several verification functions, including reverse image search, video keyframe extraction, metadata analysis, contextual analysis, image magnification, forensic filters, advanced social-media searches and video rights information. It was originally developed within the EU-funded InVID project and is particularly aimed at journalists and fact-checkers.
Why it matters for ENACT-AI
The tool is highly relevant to AI Literacy & Critical Thinking because it teaches and supports systematic verification of digital content rather than relying on intuition. It is especially useful for analysing manipulated, misleading or potentially AI-generated images and videos. The tool also strengthens understanding of digital evidence, source verification and the ways in which visual content can be taken out of context or manipulated.
Learning outcomes
Learners will be able to:
Verify the origin and context of images and videos found online.
Use reverse image search to identify previous appearances of visual content.
Extract video keyframes and use them for further investigation.
Examine image and video metadata.
Identify contextual clues that may indicate misleading or manipulated content.
Apply basic forensic filters to still images.
Compare information from different online sources before reaching a conclusion.
Develop a systematic approach to fact-checking visual content.
Understand the limitations of visual evidence and the importance of corroborating information.
Suggested activity
Give learners a potentially misleading image or short video circulating on social media. Ask them to investigate it using the InVID Verification Plugin. They should extract keyframes if it is a video, conduct reverse image searches, examine metadata where available and investigate the original context of the content. Learners then present their findings and explain which pieces of evidence allowed them to determine whether the content was authentic, manipulated, misleading or taken out of context.