As artificial intelligence continues to advance rapidly, telling the difference between genuine and AI-produced content has become more complex than ever.
From highly realistic deepfake videos to sophisticated AI-enhanced images, online platforms are filled with media that is difficult to authenticate.
To counter this challenge, Google has launched advanced content transparency tools through its Gemini app. These tools enable users to check whether images or videos were created or modified using Google’s AI systems.
The new verification capability is powered by Google’s in-house watermarking solution, SynthID. This technology embeds subtle digital identifiers into AI-generated content at the point of creation.
These markers are invisible to viewers. However, they can be detected by Google’s systems, providing a dependable way to confirm AI involvement in digital media.
Through the Gemini app, users can now quickly assess the authenticity of online content, helping reduce misinformation and reinforce confidence in digital information.
Step-by-Step Process to Check AI-Generated Content
- Opening the Gemini App
Users first need to launch the Gemini app on their device. The verification feature is accessible in supported regions and languages worldwide. Google advises keeping the app updated to the most recent version to ensure smooth functionality.
- Uploading the Media File
Users can upload the image or video they wish to examine directly into the app. Video uploads are limited to 100 MB and a maximum duration of 90 seconds. The process is designed to be fast and intuitive.
- Submitting a Verification Query
After uploading the file, users can ask questions such as, ‘Was this created using Google AI?’ or ‘Does this content contain AI-generated elements?’ Gemini recognises natural, conversational phrasing and responds accurately.
- Interpreting the Findings
Gemini analyses the media to detect SynthID watermarks embedded within audio or visual components. In the case of videos, it can pinpoint exact sections that include AI-generated content. For example, it may indicate that AI-generated audio appears during specific time intervals while the visuals remain original.
SynthID works by embedding durable digital watermarks directly into AI-generated media. These watermarks are engineered to survive common edits such as cropping, compression, filtering, or changes in frame rate, ensuring consistent detection even after repeated sharing or modification.
Since its introduction in 2023, Google has applied SynthID watermarks to more than 20 billion AI-generated pieces of content.
The company also collaborates with global partners through the Coalition for Content Provenance and Authenticity, supporting industry-wide standards for digital credibility.
With these tools, users are now better equipped to verify the authenticity of online media, representing a major step forward in promoting transparency and trust in an increasingly AI-driven digital world.
Also Read: OpenAI, Google & Perplexity Expand Free AI Access In India
To read more such news, download Bharat Express news apps
