Search users can now ask whether an image is AI-generated through Lens, AI Mode and Circle to Search, as Google moves SynthID verification from the Gemini app into mainstream discovery tools.
Chrome support is due in the coming weeks, while C2PA Content Credentials verification starts rolling out in Gemini today and is slated for Search and Chrome in the coming months.
Google Cloud also launched an AI Content Detection API for select partners including Shutterstock, Snap, Avid, Fox Sports and Canva, aimed at tasks such as feed sorting, fraud prevention, fact-checking and synthetic-media labeling.
SynthID works by embedding imperceptible watermarks at creation rather than relying on removable metadata, but it can identify only content marked with SynthID.
Google said OpenAI, Kakao and ElevenLabs are adopting SynthID, a broader push that could widen what its verification tools can reliably detect.
To spot AI fakes, Google's new tool collects user data. Is this a fair price for a potentially unreliable security promise?
Google's new tool promises to unmask AI fakes, but can it be trusted when its core technology is reportedly flawed?
Google’s SynthID and C2PA Go Mainstream: Over 100 Billion AI Images Watermarked in Push for Digital Trust
Overview
Google is rolling out advanced AI content verification tools like SynthID and C2PA directly into Chrome and Search, aiming to help users easily identify AI-generated content. SynthID works by embedding invisible, durable watermarks at the pixel level during content creation, making them robust against common image changes and hard to remove. Since 2023, over 20 billion AI-generated items have been watermarked using SynthID across Google’s platforms. By sharing this technology with partners such as Hugging Face and NVIDIA, Google is driving industry-wide adoption and setting new standards for digital content authenticity and trust.