Vertex AI by Google Introduces Invisible Watermarks

During a standing-room-only keynote address at Google Cloud Next in San Francisco, Google CEO Sundar Pichai today introduced a suite of AI tools for Google Workspace, including adding a digital watermark to AI-generated images. The tool will be a feature of Google’s Vertex AI.

Launched in 2021, Vertex AI is a part of Google’s suite of cloud computing services that lets developers train and deploy customized AI applications and models.

“Today, we are pleased to be the first cloud provider to enable digital watermarking and verification for images submitted on our platform,” Pichai told the packed auditorium. “Using technology powered by Google Deepmind, images generated by Vertex AI can be watermarked in a way that’s invisible to the human eye without damaging the image quality.”

On the other side, Pichai said, Google has developed verification tools to search for the watermarks and identify that AI created the image.

s

As generative AI improves, the ability to tell fake generated images from real images is becoming more complex. Last month, MIT researchers suggested the addition of code to AI image models that could mark an image as AI-generated.

“As exciting as the AI opportunity is, we all need to be clear-eyed about the imagerial challenges, ” Pichai said. “That’s why Google Cloud is so committed to developing and deploying technologies responsibly—with privacy, security, and safety as a goal and guided by our long-standing AI principles.”

While he did not say when the new watermark tool will be available to the public, Pichai said it is now available in an “experimental preview” to a select group of customers. 

Earlier this month, Google announced new upgrades to its popular search engine, integrating new generative AI capabilities to improve online search and understanding.