09/03/2026
šØ A DEEPER LOOK AT THE COURT RULING BEHIND MY POST šØ
First, THANK YOU for all of the engagement, shares, comments, messages, and conversations surrounding my original post. I genuinely did not expect it to reach as many people as it has, nor gain as much traction as I have. Because it has traveled so much, I think it is important that I follow it up with more context and give everyone the opportunity to understand exactly what happened in this caseā not just a shortened version of a complicated federal court ruling.
I also want to make something very clear:
I am not an attorney, and this is not legal advice. I am a photographer and business owner who read something that deeply concerned me and made a decision about how I will operate MY business going forward.
So, let me talk a little more about what actually happened. On August 25, 2026, the United States Court of Appeals for the Seventh Circuit issued its decision in United States v. Anderegg. According to the courtās opinion, Steven Anderegg was accused of using generative AI software to create hundreds of hyper-realistic sexually explicit images depicting prepubescent children. The government conceded that the images involved in the charges did not depict an actual child and could not be linked to an actual child. This distinction is incredibly important to understanding the ruling. The case began after Meta reported a potential CSAM transmission to the National Center for Missing & Exploited Childrenās CyberTipline. According to the governmentās allegations described by the court, AI-generated material had been sent through Instagram to an account belonging to a MINOR. Law enforcement eventually searched Andereggās home and electronic devices and, according to the government, discovered hundreds of AI-generated images depicting minors in sexually explicit situations. He was charged with multiple offenses, including production, distribution, transferring material to a minor under 16, and possession.
And this is where one of the most important clarifications from my original post needs to be made:
The Seventh Circuit did NOT rule that every form of AI-generated CSAM is legal.
It did NOT dismiss all of the charges against this defendant. The ruling at issue concerned the POSSESSION charge ā specifically, his private possession in his home of obscene virtual material that did not depict an actual child. The production, distribution, and transfer-related charges were NOT what the Seventh Circuit dismissed in this appeal. So why was the possession charge dismissed? The Seventh Circuit explained that it was bound by previous United States Supreme Court decisions concerning the First Amendment, obscenity, privacy inside the home, and virtual CSAM. One of those cases is Stanley v. Georgia, decided in 1969, which established constitutional protection surrounding the private possession of obscene material within someoneās home. Another extremely important case is Ashcroft v. Free Speech Coalition, decided in 2002. That case dealt with āvirtualā child po*******hy ā material that appears to depict children but was created without using actual children. The Supreme Court distinguished that material from CSAM produced through the sexual abuse and exploitation of REAL children.
Fast forward nearly 25 years.
Artificial intelligence has completely changed what āvirtualā imagery can look like. We are no longer simply talking about obviously fake drawings or primitive computer-generated images. AI can now produce imagery so realistic that distinguishing an AI-generated child from an actual child can be incredibly difficult.
And here is what I believe is one of the most important parts of this entire opinion: THE JUDGES RECOGNIZED THAT PROBLEM.
The Seventh Circuit essentially explained that although todayās technology creates serious concerns about how those older Supreme Court precedents apply, a lower federal court cannot simply rewrite Supreme Court precedent because technology has changed. The court said it was bound to follow those decisions. Two judges went even further in a separate concurrence and specifically called attention to how drastically technology has changed since the Supreme Courtās 2002 decision. They pointed to concerns that AI-generated CSAM can now be virtually indistinguishable from CSAM created through the physical abuse of an actual child. They discussed the burden AI-generated material can place on law enforcement trying to identify REAL victims. They discussed concerns surrounding AI training data. And they said additional guidance from the United States Supreme Court would be beneficial.
THAT is the part that stopped me in my tracks. This isnāt simply a headline saying, āAI-generated CSAM is legal.āThat would be an inaccurate oversimplification of what the court actually decided. The reality is much more complicated ā and, in my opinion, the complexity is exactly why we should be paying attention. Technology has advanced at a speed our laws and decades-old court precedent were never written to anticipate.
And as a photographer, I cannot ignore what that means. I photograph children. Families trust me with incredibly high-resolution photographs of their babies, toddlers, children, seniors, and families. Once an identifiable photograph of a child is publicly available online, I cannot control every screenshot, download, share, AI model, website, or person who may eventually obtain it. I cannot control how technology develops five years from now. What I CAN control is what I personally choose to put online. So my decision remains the same.
š¢ EFFECTIVE IMMEDIATELY š¢
I will no longer publicly post identifiable images of children through my photography business, signed model release or not. When I choose to share work involving children, faces may be blurred, covered, cropped, photographed in a way that does not identify the child, or otherwise made non-identifiable.
My clients will still receive their beautiful photographs.
This is about what I choose to make publicly accessible through MY business.
Will everyone agree with me?
Absolutely not.
Will other photographers make different decisions?
Absolutely.
Iām not telling another photographer how to operate their business, and Iām not telling parents they shouldnāt post pictures of their children. Iām simply explaining why I have chosen this boundary for mine.
Could it affect my engagement?
Probably.
Could it make advertising family and childrenās photography more difficult?
Possibly.
Could it cost me bookings?
Maybe. And I still donāt care.
If protecting the children who come through my doors means my social media doesnāt perform as well, thatās a trade I am willing to make every single time. I would much rather be overly cautious today than realize years from now that technology progressed in a way none of us anticipated and wish I had done more when I had the opportunity. I always want to give my honest attempt at living on the correct side of history. For me, that means adapting as technology changes ā even when the law hasnāt completely caught up yet. The safety and privacy of children and my clients will ALWAYS come before likes, shares, engagement, bookings, or my business. Whether you agreed with me or disagreed with me, people are TALKING about childrenās digital safety and the implications of rapidly advancing AI. That conversation matters. Please donāt take my word for what this court decided. Read the opinion yourself. Iāll leave the official United States Court of Appeals for the Seventh Circuit decision below so everyone can read the actual ruling and form their own opinion.
-Roberts Photographyš¤
