AI Regulation Tracker / US state legislation

Wyoming HB0102 Would Create AI Self-Harm and Deepfake Felonies

Wyoming House Bill HB0102, an engrossed bill that has not been enacted into law, would make it a felony to develop or distribute an AI system intended to promote self-harm, would add AI-generated "synthetic sexual material" of children to the sexual-exploitation offenses, and provides that using AI to commit a crime is not a defense. It would also prohibit interactive computer services from using AI to censor a user's political speech, framed as a civil trade-practices rule enforceable by the Attorney General rather than a new AI crime. If enacted, the Act would take effect July 1, 2026.

By Anthony Guerriero · The Leveraged Years Regulatory Research · Published March 7, 2026 · Last updated March 7, 2026 · State legislation

The short version

Wyoming would attach criminal and civil consequences to specific AI uses. Bottom line: HB0102, titled "Protecting kids from deepfakes and exploitative images," would put AI-generated abuse material and AI-promoted self-harm into the offense code and would prohibit AI-driven political-speech censorship by online services. Who this affects: anyone who builds or distributes AI systems used in these ways, and interactive computer services that moderate content with AI. Effective date: if enacted, July 1, 2026. What changed: the bill would create a felony for developing or distributing an AI system intended to promote self-harm, would fold "synthetic sexual material" of children into sexual-exploitation crimes, and would close the door on arguing that AI did it rather than the defendant. My read: the political-speech section is the novel one and would sit in Wyoming's trade-practices code with civil enforcement, not prison, so I keep the criminal and civil halves of this bill separate when I explain it. Status: this is an engrossed bill, not current law. Primary source: the engrossed text of HB0102.

Regulatory Briefing

Instrument: House Bill HB0102, "Protecting kids from deepfakes and exploitative images," amending Wyoming's criminal code (title 6) and adding a political-speech section to title 40
Authority: Wyoming Legislature; criminal provisions enforced by prosecutors, the political-speech provision enforced by the Wyoming Attorney General
Jurisdiction: State of Wyoming, United States (sub-national)
Status: Pending legislation. This briefing is drawn from the engrossed text of HB0102 in the 2026 session; any signing or final-passage date comes from secondary sources (LegiScan) and is not confirmed from the enrolled text
Bindingness: Proposed law, not yet in force. This is a bill. If enacted it would create felony offenses and a civilly enforceable prohibition.
Effective date: If enacted, July 1, 2026 (the effective date specified in Section 3 of the bill)
Primary Source: Wyoming Legislature, HB0102 engrossed text, wyoleg.gov
Workflow Impact: Product, trust-and-safety, and legal teams operating AI tools or AI-driven content moderation that touch Wyoming users need to map their conduct against these offenses and the political-speech prohibition.
Operational Next Step: Separate the criminal-code provisions from the political-speech provision and assess each against how your systems actually behave.
Last Verified: July 23, 2026

Start with what the bill says it would do. Its enacting clause specifies "that using artificial intelligence to commit a criminal offense shall not be a defense to the offense," and then lists the offenses it would establish, including "the use of synthetic sexual material or artificial intelligence against children" and "the use of artificial intelligence to promote self-harm or to censor political speech." That is a lot of ground, and the pieces do not all live in the same part of the bill.

The proposed criminal-code changes

Two moves would happen inside Wyoming's title 6 criminal code. First, the bill would define "synthetic sexual material" as material created, altered, or manipulated through an AI system or another digital method to realistically depict the intimate parts of a specific person, and would work that concept into the sexual-exploitation-of-children offenses. Realistic AI-generated depictions of children would be pulled squarely into those crimes.

Second, the bill would create a new self-harm offense. In its words, "unlawful development or distribution of an artificial intelligence system to promote self-harm is a felony punishable by imprisonment for not more than ten (10) years, a fine not to exceed ten thousand dollars ($10,000.00), or both." The offense would target developing or distributing a system intended to promote self-harm, with the bill setting out the intent and knowledge elements. There would be an immunity carve-out for interactive computer services and telecommunications carriers for content provided by another person, and the bill provides that a system producing such content solely as a result of a user's own prompt does not by itself establish the offense.

The no-defense rule would tie it together. A defendant could not argue that the AI, rather than the person, committed the crime. That would close an obvious gap as generative tools spread.

The political-speech provision would be civil, not criminal

The most unusual part would sit in title 40, Wyoming's trade-practices code, not the criminal code. Proposed section 40-12-801, "Prohibition of censorship of political speech by interactive computer services using artificial intelligence," provides that no interactive computer service, through an AI system, shall intentionally block, ban, remove, de-platform, demonetize, de-boost, restrict, or otherwise discriminate against a user based on the user's political speech, or modify or manipulate a user's content to censor political speech. The prohibition would apply whether the action is automated or done by hand. Because it sits in the trade-practices code, this is a civil, consumer-protection-style rule enforced by the Attorney General, not a new AI crime.

It would carve out speech that violates federal or state law, that is a true and credible threat or incitement of violence, that is obscene, or that infringes intellectual property. Enforcement would run through the Attorney General, who could bring an action and seek a civil penalty of not more than ten thousand dollars for each violation. So despite the enacting clause grouping political-speech censorship with the criminal offenses, the operative remedy here would be civil, enforced by the state, not a prison term. That distinction matters if you are assessing exposure, and it is the kind of provision likely to draw its own legal challenges given the First Amendment questions around compelling platform carriage.

Why professionals outside Wyoming should track it

If your organization ships generative AI, or moderates a platform with AI, this is a concrete example of a state moving to convert broad AI worries into specific offenses with real penalties. The children and self-harm provisions echo duties appearing in other states, so building against them once travels well. The political-speech section is more idiosyncratic and legally contested, and I would not assume it survives challenge unchanged, but it signals a second front: states legislating not only what AI must prevent but how AI-driven moderation may treat lawful speech. My practical read is to keep two separate compliance workstreams here, one for the criminal-code duties around abuse material and self-harm, which are well-grounded, and one watching the political-speech provision, which is novel and worth monitoring for litigation before you re-engineer moderation around it. The bill sets out the rule; your judgment decides how hard to lean into an untested provision, and how much weight to put on a measure that is not yet law.

Questions professionals are asking

Is the self-harm provision a crime or guidance?

It would be a crime if enacted. HB0102, an engrossed bill, provides that unlawful development or distribution of an AI system to promote self-harm would be a felony punishable by imprisonment for not more than ten years, a fine not to exceed ten thousand dollars, or both. It includes intent and knowledge elements and an immunity carve-out for services carrying third-party content.

Is AI-driven political-speech censorship treated as a crime too?

No. Although the bill's enacting clause groups it with the criminal offenses, the operative provision, proposed section 40-12-801, would sit in Wyoming's trade-practices code. It would prohibit interactive computer services from using AI to censor a user's political speech as a civil rule enforced by the Attorney General, who could seek a civil penalty of up to ten thousand dollars per violation, not a prison term.

What is "synthetic sexual material" under the bill?

It is material created, altered, or manipulated through an AI system or another digital method to realistically depict the intimate parts of a specific person. The bill would use that definition to fold AI-generated depictions of children into Wyoming's existing sexual-exploitation-of-children offenses.

Can a defendant blame the AI?

No. The bill provides that using artificial intelligence to commit a criminal offense would not be a defense to the offense, so responsibility would stay with the person, not the tool.

RELATED BRIEFINGS

Browse the full AI Regulation News tracker

Informational analysis for working professionals, not legal advice. Confirm how any statute applies to your situation with qualified professionals in the relevant jurisdiction.