When AI Creates the Image, Who Does the Law Protect?

A recent federal court ruling exposes just how unprepared our laws are for the reality of AI-generated sexual abuse.

Technology has changed.

The harm has changed.

But in many ways, our laws are still trying to answer today's questions using rules written for yesterday's technology.

On August 25, 2026, the U.S. Court of Appeals for the Seventh Circuit issued a decision in United States v. Anderegg involving sexually explicit images of children created entirely through artificial intelligence.

The ruling should get the attention of every parent, policymaker, school administrator, technology company, and person concerned about where artificial intelligence is taking us.

Because underneath the complicated constitutional arguments is a disturbing question:

If AI can create sexual images of children that look real, should the law treat those images differently simply because the child depicted does not physically exist?

What the Court Actually Ruled

The case involved a defendant accused of using generative AI to create hundreds of photorealistic sexual images depicting prepubescent children.

According to the court, the government acknowledged that the images involved in the charges did not depict actual children and could not be connected to actual children.

The defendant faced multiple federal charges, including production, distribution, transferring material to a minor, and possession.

A lower federal court dismissed the possession charge, concluding that the federal statute criminalizing his private possession of the entirely virtual material was unconstitutional as applied to him.

The federal government appealed.

On August 25, the Seventh Circuit affirmed that decision.

But there is an incredibly important distinction here:

The court did not rule that all AI-generated child sexual abuse material is protected by the First Amendment.

It addressed the specific charge involving the defendant's private possession of obscene, virtual material in his home. The other charges involving producing, distributing, and transferring the material were not eliminated by this ruling.

That distinction matters.

Still, the reasoning behind the decision should concern us.

The Problem: Our Technology Is Living in 2026. Our Precedent Isn't.

The judges weren't necessarily saying they believed this was the right outcome for the AI era.

In fact, the opinion openly acknowledged the problem.

The court explained that today's generative AI can produce images of virtual children that may be extremely difficult to distinguish from images involving actual children.

But the judges said they were bound by previous decisions from the United States Supreme Court.

One of the most important is Ashcroft v. Free Speech Coalition, decided in 2002.

Think about that.

Before Instagram.

Before TikTok.

Before today's smartphones.

And more than two decades before ordinary people could type a few words into an AI program and generate photorealistic sexual imagery.

The Seventh Circuit acknowledged how dramatically technology has changed, but explained that lower federal courts cannot simply rewrite Supreme Court precedent because technology has advanced.

Two judges went even further in a separate concurrence, saying the Supreme Court could provide additional guidance about how the First Amendment should apply to virtual CSAM in the age of modern artificial intelligence.

That should be a wake-up call.

This Is Where My Story Becomes Part of This Conversation

This issue isn't theoretical to me.

When I was 14 years old, photos and videos were taken from my private social media accounts without my permission.

Artificial intelligence was used to manipulate my images and create sexually explicit deepfakes.

The images weren't real.

I was.

The body created or manipulated by technology didn't have to physically exist for the humiliation to exist.

It didn't have to be real for people to see it.

It didn't have to be real for people to talk about it.

It didn't have to be real for a child to have to walk into school knowing that sexually explicit images carrying her face were being shared.

And that is one of the biggest problems I see as courts, lawmakers, and technology companies attempt to deal with AI-generated sexual material.

We keep asking whether the image is real.

We also need to ask whether the harm is real.

There Is an Important Difference in My Case

I also want to be careful not to pretend that my experience and United States v. Anderegg are legally identical.

They aren't.

The images discussed in this particular possession charge did not depict an identifiable real child.

My experience involved my actual likeness.

My real photographs.

My real face.

My identity.

That distinction is incredibly important legally.

But these situations belong in the same larger conversation because AI is rapidly destroying the boundaries our laws have traditionally relied upon.

What is real?

What is fake?

Who is the victim?

What happens when a fake body is attached to a real child's face?

What happens when an entirely synthetic image is indistinguishable from an image created through the abuse of a real child?

And what happens when investigators can't immediately tell the difference?

These aren't questions for the future.

They are questions we have to answer now.

We Cannot Wait Another 20 Years

I respect the role of our courts and the constitutional protections they are required to uphold.

But this decision demonstrates why courts cannot solve this problem alone.

Congress needs to act.

State legislatures need to act.

Technology companies need to take responsibility for what their tools make possible.

Schools need policies specifically addressing AI-generated sexual images.

And our laws need to recognize the different forms of harm created by this technology—including situations involving entirely synthetic content and situations where the face, voice, body, or identity of a real person is manipulated.

We also need laws written carefully enough to survive constitutional challenges.

Because passing a law isn't enough.

It has to actually work when it reaches a courtroom.

"No Real Child" Cannot End the Conversation

There is something deeply uncomfortable about discussing AI-generated sexual images of children and hearing the phrase:

"No actual child was involved."

Sometimes that may be an important legal distinction.

But it cannot be the end of the conversation.

AI is becoming capable of producing imagery that can be extraordinarily realistic.

And when a real child's likeness is used, there absolutely is a real person on the other side of that screen.

I know because I was one of them.

Artificial intelligence has advanced faster than our legal system.

Now we have a choice.

We can wait for more victims, more court cases, and more technology to expose the holes in our laws.

Or we can start fixing them.

My story was stolen. My voice wasn't.

And I will continue using that voice to push for laws that understand something technology should never allow us to forget:

Children deserve protection whether the weapon used against them is physical or digital.

Voices Unstolen
Real stories. Real strength. Real hope.