Media Party New York

It’s becoming more difficult to tell the difference between an AI generated image and a real photo just by looking at it. Deepfake detection tools have been used in the past, but with detection you are playing from behind.  Not only in keeping up with the detection technology, but also because the content has often already been distributed. In some cases, like the AI generated image of an explosion at the Pentagon, real damage is done long before detection results can be confirmed and disseminated. A more proactive way to combat the threat of deepfakes is to make sure the content you are creating, distributing, and consuming is traceable to its origin. Content Credentials can clearly, and securely indicate to a consumer whether an asset was created with a camera or AI. Content Credentials are a new kind of secure, tamper evident metadata added to digital assets that reveal its provenance including its origin. They show what edits have occurred since its origin, and communicate attribution information and how it was published. In this workshop, Tim Murphy, from Pixelstream, will demonstrate how Content Credentials work and how your organization could leverage this technology to protect your assets, establish trust with your readers, and make deepfakes a thing of the past. We will use WordPress as an example platform that this can be deployed on, but it is applicable to most other website environments as well.