How to verify AI-generated images and videos with Google’s Gemini tool

As AI-generated images and videos spread rapidly across social media and messaging platforms, distinguishing real content from manipulated media has become increasingly challenging. While visible watermarks offer one clue, many AI-generated visuals now carry invisible identifiers that are harder to detect. Google is now offering users a practical way to verify such content through its Gemini AI model, which can identify hidden AI watermarks embedded in media files, according to company information.

Google has made Gemini available across products such as Gmail and Docs, and is positioning the tool as a way to help users differentiate between authentic and AI-generated content using the latest watermarking technologies adopted across the industry. The company said Gemini can analyse uploaded images and videos to detect SynthID, a digital watermark designed specifically for AI-generated media.

To check whether a photo or video is AI-generated, users need to open the Gemini app on their phone and upload the media file they want to verify. Google informed that video uploads must not exceed 100MB in size and should be no longer than 90 seconds. Once uploaded, users can ask Gemini to verify the content by entering prompts such as whether the media was generated using Google AI or whether it is AI-generated using any tool.

Gemini then scans the content for the presence of the SynthID watermark, which can be embedded in both the audio and visual layers of AI-generated media. Google stated that Gemini has been trained to analyse these markers and identify which parts of the video or image were generated using AI, along with the duration for which AI-generated elements appear.

According to Google, SynthID watermarks are designed to be detectable only by AI systems and related tools, making them invisible to the human eye and difficult to remove without degrading the content. The company launched SynthID in 2023 and informed that more than 20 billion pieces of AI-generated content have since been watermarked using the technology.

Google further stated that it is working with third-party companies to extend SynthID watermarking to content generated by other AI tools, with the aim of improving transparency and enabling more reliable identification of AI-manipulated media across platforms.