Why You Won't See Children's Faces on My Social Media Anymore

I've taken down all photos of children where faces are identifiable from my social media accounts. Going forward, I won't be posting them either.

Client galleries - which will now always be password protected - will still feature your kid's smiling faces, don't worry!

If you aren't aware, a recent court decision (United States v. Anderegg from the 7th Court) ruled that AI-generated child sexual abuse material is protected by the First Amendment. Federal judges apparently felt constrained by a 20-year old Supreme Court precedent (Ashcroft v. Free Speech Coalition.

In the case, a man in Wisconsin used generative AI to produce thousands of realistic, explicit images of "virtual" children. While this ruling only recognizes the ownership as free speech, he still faces prosecution for production and distribution - which I personally hope he is found guilty for.

For those who don't know exactly how generative AI works, in a nutshell an AI model is first trained on information first. Any and all publicly accessible data is fed into a complex matrix of computing, instructions, and code and out pops a predictive machine.

That training data, where do you think that came from? The internet, of course, and where can you find the most available source of publicly available images in the world? Social media! And just like authors and artists who have had their works fed into AI without their permission, all those photos we've posted in the past were unwillingly and/or unknowingly used in those models.

While maybe not the intended result, that training data is what made the AI generation of child sexual content possible. So in my mind, even if the generated content was not a real person, every AI generated image of a virtual person is based on some combination of images of real people.

Follow that path and you get this: social media images of children eventually led to the creation of AI-generated child sexual content.

That's just not something that I can get behind. So I've stopped.

I completely understand there are some contradictions or hypocrisies to this decision:

• I'm not removing photos from my website

• All the data had already been fed / will continue to be available

• One photographer's images being removed from the machine is not going to produce meaningful change

• I continue to use AI in other ways, which supports the very technology and processes used to create the offending content

From where I stand though, this is more a statement and a promise to not be a part of what feeds the ability to generate this content. I can't stop AI from being abused, but I can stop supplying fuel for some of the horrendous results that can come from the tech.

When it came down to it, this is what I felt was right as a father and world-citizen who is responsible for protecting kids who can't protect themselves.