Google has taken a giant leap towards ensuring the authenticity of AI-generated content.
The tech giant's SynthID watermarking tool, initially designed for AI-generated images and audio, now extends its capabilities to videos and texts as well.
Demis Hassabis, CEO of Google's DeepMind, unveiled the upgraded SynthID system at the recent Google I/O developer conference. The system can now imprint watermarks on digitally generated videos and AI-created texts. This is a significant development as the misuse of AI technology for spreading misinformation and creating nonconsensual content has been a growing concern.
Want to learn more about AI's impact on the world in general and property in particular? Join us on our next Webinar! Click here to register
SynthID, announced last August, imprints AI imagery in a way that is invisible to the human eye but detectable by the system. This approach differs from other watermarking protocols like C2PA, which adds cryptographic metadata to AI-generated content.
In addition to images, SynthID can also inject inaudible watermarks into AI-generated music created using DeepMind's Lyria model. This expansion of SynthID's capabilities is part of Google's broader effort to develop safeguards against the misuse of AI technology. The Biden administration has also directed federal agencies to build guidelines around such safeguards.
As AI technology continues to evolve, tools like SynthID are crucial in maintaining the integrity and authenticity of AI-generated content. By expanding its capabilities, Google is making a significant contribution to the fight against the misuse of AI.
Want to learn more about AI's impact on the world in general and property in particular? Join us on our next Webinar! Click here to register
Made with TRUST_AI - see the Charter: https://www.modelprop.co.uk/trust-ai
Comments