Google Gemini Can Now Detect AI-Generated Images and Videos via SynthID

Google has added AI content verification to its Gemini app, allowing users to check whether images or videos were created or edited using Google's AI tools. The feature relies on SynthID, Google's invisible watermarking technology, which Gemini scans across both visual and audio tracks in videos to identify AI-generated segments. Users can simply upload a file and ask a plain-language question, such as 'Was this generated using Google AI?' to receive an attribution-based result. Gemini also supports Content Credentials, a separate digital provenance mechanism that records a file's origin history. The feature supports files up to 100 MB for images and videos up to 90 seconds, is currently available on Android, and is limited to detecting Google's own AI watermarks rather than serving as a universal AI-content detector.