The Conversational AI Services Act establishes safety, transparency, and parental supervision requirements for AI services interacting with minors and mandates FTC enforcement to prevent deceptive practices.
Mariannette Miller-Meeks
Representative
IA-1
The Conversational AI Services Act establishes essential safety and transparency standards for AI services interacting with minors, including mandatory disclosures, parental control tools, and protections against harmful content. The bill also prohibits deceptive practices, such as AI masquerading as human or providing unauthorized mental health advice, and mandates protocols for suicide and self-harm prevention. Enforcement of these requirements is granted to the Federal Trade Commission to ensure compliance and consumer protection.
Conversational AI Services Act Mandates Age-Appropriate Guardrails: 3-Hour Disclosure Limits and Crisis Protocols Set to Launch
The Conversational AI Services Act is a new rulebook for the chatbots and digital personas our kids are increasingly hanging out with. It targets 'conversational AI'—think of the apps your teenager uses to role-play or the AI assistants your middle-schooler asks for homework help—and forces them to follow strict safety and transparency standards. The bill requires these services to tell users under 18 that they are talking to a computer, not a human, at least once every three hours of continuous use. It also bans 'variable reward' systems, which are those unpredictable points or badges designed to keep kids hooked on an app like a digital slot machine.
Setting Digital Boundaries
Under Section 2, the bill tackles the 'uncanny valley' of AI-human relationships. It requires companies to implement 'reasonable measures' to prevent their AI from acting as a romantic partner or sexual interest for minors. This isn't just about blocking bad words; the law specifically prohibits AI from simulating emotional dependence or role-playing adult-minor romantic relationships. For a parent, this means the AI your child interacts with is legally barred from 'love-bombing' them or pretending to have human feelings. Additionally, the bill mandates that these services provide tools so parents can actually see and manage their kids' screen time and account settings, moving the needle from 'wild west' to a supervised digital playground.
Mental Health and Crisis Support
One of the most practical shifts involves how AI handles heavy topics. If a user mentions suicide or self-harm, the bill requires the operator to have a protocol in place to immediately refer that person to a crisis service, like a hotline or text line. It also draws a hard line on medical advice: AI cannot pretend to be a licensed psychologist or behavioral health professional. For someone struggling with their mental health, this ensures the AI acts as a bridge to real-world help rather than a substitute for a doctor. While the bill uses the term 'reasonable measures'—which gives companies some wiggle room on how they build these filters—the Federal Trade Commission (FTC) is given the teeth to fine companies that don't take these safety duties seriously.
What This Means for the Tech Industry
If you develop or run a chatbot, the clock starts ticking soon. The Act takes effect one year after it is signed, giving companies 12 months to overhaul their code and user interfaces. While big tech players likely have the resources to comply, smaller startups might find the 'three-hour disclosure' rule and the complex crisis-referral protocols a heavy lift for their engineering teams. The bill exempts internal business tools and simple customer service bots (like the one that helps you track a package), focusing its energy on the general-purpose AI that aims to simulate human connection. By defining 'minors' as anyone the operator is 'reasonably certain' is under 18, the bill also signals that companies can't just look the other way when it comes to the age of their users.