Iowa lawmakers took a significant step during the 2026 legislative session in addressing concerns about artificial intelligence and child safety by approving new safeguards for AI-powered chatbots that interact with minors.
The measure, signed into law by Gov. Kim Reynolds last month, requires chatbot operators to disclose that users are interacting with artificial intelligence, establish parental controls and implement protections for children who express suicidal thoughts or self-harm.
But while prosecutors and law enforcement officials praised the legislation as an important first step, they say Iowa remains vulnerable to another rapidly growing threat: AI-generated child sexual abuse material.
County attorneys and sheriffs across Iowa are urging lawmakers to revisit legislation next year that would criminalize AI-generated child sexual abuse images that do not depict actual children but are created entirely through artificial intelligence. They argue current state law leaves a gap that allows the possession, creation and distribution of such material.
Mahaska County Attorney Andrew Ritland, who helped draft legislation targeting AI-generated child sexual abuse material, said the technology is already being used to facilitate child exploitation and sextortion schemes.
“We're also nationally seeing the same problem with kind of sextortion, so a person will get a genuine image or try to create a fake image and use that to basically compel a child to produce genuine child sex use material,” Ritland said during a recent interview. “There are very real harms, even though they're wholly generated, they're very much real harms with this material being out there.”
Dubuque County Sheriff Joe Kennedy said his office investigated a case last year involving male high school students who created and shared images depicting female classmates' faces attached to artificially generated nude bodies. The investigation required extensive documentation, preservation requests, subpoenas and search warrants, and was complicated by the fact that some AI platforms are operated by companies outside the United States.
“When you have an app that’s based out of Canada, it’s totally different,” Kennedy said. “They don't have to comply with our subpoenas.”
The case underscored how quickly the technology is spreading among young people and how difficult such investigations can be.
Background
Lawmakers have approved several measures in recent years aimed at addressing emerging harms posed by artificial intelligence.
In 2024, Reynolds signed Senate File 2243, which criminalized the creation, adaptation or modification of images of real children to make them appear engaged in prohibited sexual acts. The law expanded Iowa's definition of child sexual abuse material to include images that have been altered to make an identifiable minor appear to be engaged in sexual activity. However, the law applies only when an actual child is depicted and does not address images generated entirely by artificial intelligence.
This year, lawmakers unanimously passed and Reynolds signed into law Senate File 2417, a measure establishing requirements for AI-powered conversational chatbots that interact with minors.
The legislation requires chatbot operators to clearly disclose that users are interacting with artificial intelligence, provide parental controls and privacy management tools, and implement safeguards designed to prevent sexually explicit interactions with minors. It also requires operators to adopt protocols for responding to suicidal ideation or self-harm. And it prohibits chatbots from representing themselves as licensed mental health professionals.
The measure passed both chambers unanimously.
The law was prompted by growing concerns nationwide about harmful interactions between children and increasingly sophisticated AI systems.
Ritland, the Mahaska County Attorney, said the chatbot legislation is “a good start” but argued additional safeguards and enforcement mechanisms may ultimately be necessary.
“It makes sure that these chat bots aren’t trying to deceive a child, (by portraying) that they're a real person,” Ritland said. " … So it is a good start. However, it shouldn’t be the last word on the topic."
Kennedy likewise described the measure as a positive first step while warning that lawmakers face a constant challenge keeping pace with rapidly evolving technology.
What’s happened since
Attention has increasingly shifted toward AI-generated child sexual abuse images that do not involve identifiable children but can still be used to facilitate exploitation.
Ritland worked with Rep. Helena Hayes, R-New Sharon, on House File 826, which would classify AI-generated child sexual abuse images as obscene material involving a minor. The bill unanimously passed the Iowa House in 2025 but stalled in the Senate during the 2026 session.
The proposal would prohibit the production, distribution, receipt or possession with intent to distribute obscene visual depictions of minors engaged in sexually explicit conduct. The legislation specifically provides that a violation does not require that the depicted child actually exist, allowing prosecutors to pursue cases involving AI-generated images.
The bill was drafted to mirror federal law enacted through the PROTECT Act of 2003, which makes it illegal to produce, distribute, receive or possess with intent to distribute obscene depictions of artificial minors engaged in sexually explicit conduct. Ritland said Iowa lacks a comparable state-level statute, leaving many cases dependent on federal authorities that typically focus on larger, multistate investigations.
Under the proposal, first-time offenders would face an aggravated misdemeanor, while second and subsequent offenses would be punishable as Class D felonies. Convicted offenders would also be required to register as sex offenders for 10 years. The legislation includes exemptions for bona fide law enforcement and judicial activities, as well as educational materials used in accredited schools, public libraries and educational programs.
Ritland said the bill closes a critical gap in Iowa law by bringing state statutes in line with existing federal standards while targeting only obscene visual depictions and preserving constitutional protections.
Hayes said the legislation is needed because of the rapid advancement of artificial intelligence technology and the “threats that that poses to the very vulnerable in our communities, especially small children and minors,” she said during a House Judiciary Subcommittee hearing last year.
CSAM terminology
Ritland said his office and prosecutors around Iowa are increasingly encountering situations in which AI-generated images are mixed with traditional child sexual abuse material, making investigations and prosecutions more difficult.
“As technology and AI has progressed, there is effectively no real way to tell between the two,” he said. “It's complicating the prosecution of traditional CSAM (child sexual abuse material) with this AI general material being out there.”
CSAM refers to visual content depicting the sexual abuse or exploitation of children, including photographs, videos, livestreams and, increasingly, AI-generated images.
Law enforcement agencies and child protection organizations have largely adopted the term CSAM in place of the older phrase “child pornography,” arguing that the previous terminology can imply consent and fails to accurately describe the abuse depicted in the material.
Ritland said current Iowa law generally does not prohibit fully AI-generated child sexual abuse material unless it is distributed to a minor — creating a legal gap that leaves many forms of AI-generated child sexual abuse material beyond the reach of state prosecutors.
He also warned that AI-generated CSAM poses unique challenges because many images are created using artificial intelligence systems trained on vast quantities of content available online, including existing child sexual abuse material. Supporters of House File 826 argued that even when an AI-generated image does not depict a real child, the material can normalize sexual exploitation of children and encourage offenders to seek out or produce images involving actual victims.
Ritland warned that offenders use AI-created images while posing as minors to solicit real sexual images from children and to extort them in sextortion schemes.
“AI CSAM inspires individuals to engage in the sexual abuse of minors, serves as a tool for offenders to groom and seduce children online, supports the market for traditional CSAM and complicates the prosecution of existing child exploitation cases,” Ritland said in a legislative handout supporting House File 826.
Warning students
Dubuque County Sheriff Kennedy said law enforcement agencies are focusing on education as they work to prevent future incidents involving AI-generated sexual images and online exploitation.
Following last year's deepfake investigation, the Dubuque County Sheriff's Office, Dubuque Police Department and Dubuque County Attorney's Office developed a presentation for students called “Think Before You Share!”
The presentation warns students that receiving, possessing or sharing sexual images of minors can carry serious legal consequences, even when a student did not originally create the image. Law enforcement officials instruct students who receive a nude image of a minor to immediately report it to police, school officials or a trusted parent rather than forwarding it to others. Once law enforcement has been notified, students are told to follow investigators' instructions and avoid retaining or sharing the material.
The presentation also emphasizes that social media platforms and digital services often preserve user data even when users believe it has been deleted. According to the training materials, law enforcement can obtain records from platforms such as Facebook, Instagram, Snapchat, BeReal and TikTok, as well as data stored on cellphones, tablets and computers, through subpoenas, search warrants and other investigative tools.
Students are also warned that Iowa law criminalizes a wide range of conduct involving child sexual abuse material, including creating, possessing, promoting or distributing images depicting minors engaged in prohibited sexual acts.
Depending on the offense, penalties can range from misdemeanors to felony convictions carrying prison sentences and sex offender registration requirements. The presentation further warns that sharing sexually explicit images of another person without consent, including AI-generated images, can result in criminal charges.
Kennedy said the goal is not to frighten students but to help them understand the consequences of misusing rapidly evolving technology. The presentations also encourage conversations between students and parents about digital responsibility, privacy and the real-world harm caused when intimate images are created or shared without consent.
What’s next
Both Kennedy and Ritland said lawmakers should revisit House File 826 when the Legislature reconvenes next year.
Ritland said he plans to continue advocating for the proposal and will seek to get it reintroduced next session. He argued that AI-generated child sexual abuse material will become increasingly realistic and widespread without additional legal safeguards.
Kennedy said stronger penalties and clearer statutory definitions are needed before AI-generated child exploitation cases become more widespread.
“We need to get cleaner and more defined,” Kennedy said. “ … Otherwise, I mean, this is just going to get out of control.”
