The Adobe symbol that encodes AI-generated content

In the sea of ​​contents generated by artificial intelligence and considering the emergence of numerous concerns and disputes regarding the topic of authenticity, Adobein collaboration with other leading technology companies, has introduced a symbol designed to encourageAI-generated content labeling.

What the US software house calls “transparency icon” is ready to revolutionize the way we perceive and consume media created with the assistance of AI, because it will provide insight into ownership, the AI ​​software used in the genesis of the content, along with other production details. Beyond that, it ensures the fair recognition to the work of the creators.

C2PA’s “transparency icon” is pioneering

The symbol uses the initials “CR” which they stand for Content CRedentialsto avoid being confused with the Creative Commons (CC) icon, and is the result of collaboration between Adobe, Microsoft, Nikon, Leica, Camera Bits, Truepic and Publicis Groupe.

Collectively, they form the Coalition for Content Provenance and Authenticity (C2PA)a group dedicated to defining technical standards for certifying the origin and provenance of digital content.

Operationally, content credentials can be embedded in the metadata of images, videos and PDFs via the photo and video editing platforms of Adobe come Photoshop or Premiere and possibly Bing Image Generator di Microsoft. This ensures that even the most skilled manipulators cannot erase them.

For viewers, when they encounter online content marked with the acronym “CR”, they can click on the placeholder identify the rightful ownerfind information on the AI ​​tools used in the creation of the content, as well as insights into the production process.

At the same time, C2PA released a verification functionwhereby users can upload an image labeled CR and view theentire history of changes of that image, up to the point where the symbol was assigned.

The responsibility of AI is delegated to humans

Andy ParsonsSenior Director of Adobe’s Content Authenticity Initiative, describes the symbol like a digital “nutrition label”., which offers viewers vital information about media disseminated online. Its introduction is in line with the industry’s push towards transparency of AI-generated content and it relies on human responsibilityrequiring individuals to consciously incorporate it into their content.

Although different organizations have previously used different methods, this symbol aims to unify and simplify the tagging process of the data generated by AI. For this reason, in the coming months other C2PA companies will begin to implement the new symbol. Even Microsoftwhich has adopted a custom digital watermark in content created with the Bing Image Generator, will implement the “CR” iconfurther accelerating its adoption.

Content CRedentials for authenticity in the age of deepfakes

To renew the request to identify a standard method for proving authenticity It was above all the rise of AI-generated content. In particular, the threat of deepfake images and videos used in contexts such as advertising campaigns, has prompted politicians and regulators to develop proposals to prevent their use. So much so that a number of tech companies, including Adobe, have signed a non-binding agreement with the White House to develop watermarking systems to identify content generated by AI.

Demonstrating the industry’s commitment, SynthID by Google which identifies an image generated by AI thanks to a watermark imperceptible to the human eye, is another example of a tool designed to ensure responsible media consumption. With any luck, the AI ​​symbol could ripple through future iterations of the internet, reminiscent of the way the Creative Commons tag gained widespread popularity in the early 2000s.

to know more: Computer security: guide to safe browsing on the web

Leave a Reply

Your email address will not be published. Required fields are marked *