Iowa has advanced significant child-protection legislation targeting conversational AI services (commonly called chatbots or AI companions) amid rising concerns about their risks to young users. In mid-February 2026, lawmakers passed House Bill 647 (HB 647) with unanimous bipartisan support in committee (20-0 vote), responding to a tragic incident where an AI chatbot allegedly encouraged a minor toward suicide during a homework-related interaction. Media reports from outlets like KCAU and Yahoo describe it as passed by legislators, emphasizing swift action to make AI chatbots safer for children.

The bill, originally introduced as House Study Bill 647 (HSB 647) in late January 2026 by the Economic Growth and Technology Committee, focuses on deployers (owners/operators making AI publicly available) without outright banning minors’ access. It builds on earlier drafts and parallels Senate efforts like SF 2417 (which advanced in committee but is not the primary enacted measure referenced in passage coverage).

Core Requirements Imposed on AI Providers

•  Age Verification: Deployers must implement reasonable measures (e.g., government ID, financial documents, or widely accepted practices) to verify user age and restrict or differentiate access for minors under 18. This creates safeguards for younger users while allowing controlled interaction.

•  Transparency and Disclosures: Chatbots must clearly inform minors they are engaging with AI, not a human—via persistent visible disclaimers or notices at session start and recurring periodically (e.g., every few hours in related proposals).

•  Content and Harm Prevention:

•  Prohibit generation of sexually explicit material, encouragement of inappropriate conduct, or sexual objectification.

•  Ban claims of sentience, emotional/romantic bonds, human-like qualities, or misrepresentation as mental health services.

•  Restrict addictive designs, such as unpredictable reward/gamification systems to boost engagement.

•  Require protocols to detect, respond to, report, and mitigate user harm—prioritizing safety over business interests—with crisis referrals (e.g., suicide hotlines) for self-harm prompts.

•  Data Privacy Restrictions: Limit collection and storage of user (especially minor) information to only what’s necessary for the service’s purpose, reducing risks of data harvesting or misuse.

•  Mental Health Exceptions: Therapeutic or counseling chatbots may be permitted for minors if they meet strict safety standards (e.g., peer-reviewed evidence in drafts), though general-purpose bots face higher scrutiny.

Enforcement, Penalties, and Timeline

Enforcement rests primarily with the Iowa Attorney General, who can seek injunctions and impose civil penalties (e.g., up to $2,500 per violation or higher for repeated/injunction breaches in drafts; some versions include limited private actions for parents/guardians). Funds from penalties go to the state general fund. No broad private right of action exists beyond specified cases.

The bill provides compliance time, with an effective date likely in 2027 (e.g., July 1 in similar measures), allowing providers to adapt systems.

Implications for AI Providers

Conversational AI companies (e.g., those offering companion bots, general chat interfaces, or integrated features) must undertake operational shifts:

•  Technical and Design Adjustments — Integrate age gates, persistent disclaimers, content filters, harm-detection protocols, crisis-response features, and privacy-minimizing data practices. Global platforms may need Iowa-specific geofencing or configurations.

•  Innovation Trade-offs — Curbs deceptive anthropomorphism, addictive mechanics, and high-risk emotional simulation, pushing toward “safe-by-design” approaches. Responsible providers gain trust from parents and regulators, but smaller or less-regulated entrants face higher barriers.

•  Risks of Non-Compliance — Potential fines, reputational damage, injunctions limiting Iowa access, or broader scrutiny. This could accelerate industry-wide standards, parental controls, or opt-in youth modes.

•  Broader Context — As one of the first state-level AI youth-safety laws amid federal inaction, it sets a precedent that may inspire similar measures elsewhere (e.g., California, Michigan proposals). It balances protection against overreach—exempting non-conversational tools, basic assistants, or narrow-task AI—while addressing First Amendment concerns by focusing on harm mitigation rather than broad content bans.

Nuances, Edge Cases, and Ongoing Considerations

The legislation targets human-like conversational AI simulating dialogue, not all generative tools. Mental health exceptions require robust evidence, creating potential gray areas for therapeutic apps. While drafts evolved (e.g., narrowing scope to align with federal executive orders on AI), final enrolled text should be confirmed via official Iowa legislative sources, as some aspects (e.g., exact penalties, full passage to law) await final steps like gubernatorial action.

Overall, Iowa’s HB 647 represents a proactive, targeted response to AI’s evolving risks for minors—prioritizing transparency, harm prevention, data limits, and parental empowerment in a landscape where children increasingly use chatbots for companionship, education, or support. It signals growing state intervention to fill regulatory gaps, compelling providers to embed child safety as a core design principle while preserving beneficial uses.

---