Google Gemini Is Getting Better at Identifying AI Fakes

Why It Matters

By embedding verifiable AI provenance directly into content, Google is addressing growing concerns over deepfakes and misinformation, and its push for industry‑standard C2PA credentials could set a baseline for automated detection across platforms, influencing how the media ecosystem handles AI‑generated material.

Summary

Google announced that Gemini can now identify whether an image was created or edited by a Google AI tool, letting users ask “Is this AI-generated?” within the app. The feature initially supports only images using Google’s SynthID watermark, but Google says video and audio detection will arrive soon and plans to integrate verification into Search. Google also revealed that images from its new Nano Banana Pro model will carry C2PA metadata, and it aims to expand support for industry‑wide C2PA content credentials, enabling detection of AI‑generated media from other tools such as OpenAI’s Sora. The move follows TikTok’s adoption of C2PA metadata for its own AI‑generated content watermarking.