There have been great changes recently and AI systems have continued their rapid evolution. Not only are new Generative AI systems making complex tasks much easier for the average user, but the systems themselves are also becoming easier to use as they are integrated into familiar products. This talk will look at the state of the moment, including the benefits risks and challenges to law enforcement, children and the public at large.

Developments have progressed from static image generation to live video. Integration into software like Adobe Photoshop means a user can make quick adjustments to existing images, enabling law enforcement to quickly sanitize an image, a missing child poster to be rapidly made media ready, or an unwanted object removed by a novice user. Other tools allow for complex manipulation of video. Simplification of working with audio. Even the automated identification, transcription or translation of speech.

With these changes has come the need for technologies to identify generated media. Efforts like the Content Authenticity Initiative have set out to make it easier for public identification of manipulated materials. These tools, like Gen AI tools themselves are expected to see rapid development.

These technologies don't come without risks and challenges which will also be covered. Preparing law enforcement for future challenges will be key to making sure they are able to effectively understand what offenders are capable of. Victim identification efforts are sure to be put under additional load as determining whether a depicted child is real, partially real or entirely generated because a regular part of the workflow.

The end goal of this talk is to provide a look at the state of the art and prepare the audience for the Threats and Opportunities ahead.