Content Authenticity Initiative

And Adobe, the software manufacturer, called and we started looking at some stuff not related to AI around business development.

But eventually, Adobe launched something called the Content Authenticity Initiative, which is a global community of individuals and companies and technologists and others really focused on trying to establish the provenance or the origins of digital files.

Where do they come from? How might they have been manipulated? And making some or all of that information visible to the viewer.

So you can think about it like a digital nutrition label in the same way that you would go into a supermarket and pick up a can of beans and quickly determine many things about the content, where it came from, etc.

Ähm, äh, äh, äh, bin ich jetzt doof?

C-C-CAI Content irgendwas.

Content, äh, ist egal.

Auf jeden Fall gibt es da ja mehrere, gibt es da ja mehrere Systeme.

Jaron Schneider at PetaPixel finally explains why.

The big push has been the Content Authenticity Initiative and their C2PA standard, which nearly every major camera and smartphone brand has adopted.

And Apple just wasn't there.

Turns out they were quietly building their own thing, Apple Reference Image.

So it seems inevitable that camera and smartphone manufacturers will need to in- adopt some form of authenticity verification to combat the onslaught of realistic-looking AI content that is flooding all of our screens these days.

For a few years now, the Content Authenticity Initiative has been trying to roll out something like this using C2PA technology, which essentially creates a record of what has been done to an image by providing it with timestamps and content credentials.

While this solution has been broadly agreed to by most camera manufacturers, the actual rollout to consumers hasn't materialized, and it doesn't seem like there's much going in the way of real implemen- implementation of this technology yet.

Recently, however, Apple has unveiled its own plan for managing image authenticity called Apple Reference Image, which digitally signs the image at the moment the light hits the sensor, and thereby they're arguing that this offers a more secure approach to authentication that can't e- be easily faked.