Thanks in part to tools like Nano Banana, AI-generated content is more unavoidable than ever. As announced at I/O, Google is doing its part to help curtail the situation it’s played a part in, building on top of SynthID with some new AI verification checks.

Google says its SynthID verification tools have been used 50 million times globally since the ability to check images, video, and audio was added to Gemini, and now it’s bringing that same functionality to Search and Chrome. Starting today, Lens, AI Mode, and Circle to Search will be able to tell you about an images origins just by asking if a specific photo was made with AI. Just drop your photo in, and Gemini is able to verify the content shared using built-in metadata, including camera information, further AI-based editing tools, and more. Chrome users will see the same toolset roll out over the coming weeks.

As usual, everything here revolves C2PA Content Credentials, an industry standard Google has been leaning on for a while now. On-device C2PA support first arrived on the Pixel 10 series, meaning every photo taken within the native camera app included these details to allow for AI verification, and Google is bringing support for that technology to video capture on the Pixel 8 series and later. SynthID also includes new partners in Nvidia, OpenAI, Kakao, and ElevenLabs, so it’s not just your Pixel photos that will be capable of tracing AI involvement.

Google also used this opportunity to announce new partnerships with creation companies across the industry. As part of Google Cloud’s Gemini Enterprise Agent Platform, businesses will be able to catch AI content made from various models, including Google’s, to have better knowledge on what they’re publishing to their respective websites or social media accounts. Listed partners include Shutterstock, Avid, and Canva, with Adobe serving as a pretty notable absence.

Finally, building on Google’s Pixel-friendly relationship with Meta, authentic photos and videos posted to Instagram from Pixel will now be recognized and labeled as such, allowing for creators to avoid genAI accusations from their followers.