On July 26, 2026, during an Ankeny, Iowa roundtable, Senator Ted Cruz and Representative Ashley Hinson laid the groundwork for a new federal push to require parent-managed accounts for children who use AI chatbots — moving the online child safety debate beyond social media feeds and directly onto the type of AI tools already embedded in Windows, Edge, and everyday apps. The proposed legislation, known as the CHATBOT Act, would give parents oversight of their kids’ interactions with systems like ChatGPT and Claude, with mandatory family accounts for users under 13 and default safety settings for teens.

What the CHATBOT Act Proposes

The bipartisan CHATBOT Act — formally the Children’s Health, Advancement, Trust, Boundaries, and Oversight in Technology Act — was introduced in April 2026 by Senator Cruz and Senator Brian Schatz, a Democrat from Hawaii. According to Schatz’s announcement of the bill, it aims to put parents “in charge of how children access and use covered AI chatbots” rather than leaving those decisions to platform operators.

Under the bill, covered AI services would have to:

  • Create family accounts for children. These are mandatory for users under 13 and optional for teens.
  • Enable protective safety settings by default for minors.
  • Classify users as minors or adults through age-verification processes.
  • Obtain parental consent before letting a child use a chatbot.
  • Prohibit targeted advertising to children using the service.
  • Limit manipulative design features that might keep children engaged longer than intended.

The legislation does not outright ban AI chatbots for kids but instead erects a structure where a child’s access level, privacy, and usage are tied to an adult-managed account. This is a departure from earlier youth-safety bills that focused almost entirely on social media platforms.

Why AI Chatbots Are the New Target

Representative Hinson called AI chatbots the “new frontier” in child online safety during the roundtable, a phrase that captures why this policy shift is happening now. Unlike social media feeds — where content is posted by others and amplified by algorithms — a chatbot generates a personalized, conversational response to each prompt. It can sound warm, human, and authoritative, blurring the line between a tool and a confidant.

“We want to make sure that the chatbots are also in their own lane,” Hinson said, referencing the danger of an AI system “masquerading as a professional” or offering medical or mental-health advice to a minor. The House-passed Kids Internet and Digital Safety Initiative, which she cited, already contains provisions that would require AI chatbots to avoid impersonating licensed experts and to provide crisis resources when a child signals self-harm or suicide.

Cruz, for his part, framed the CHATBOT Act as an extension of the Kids Off Social Media Act (KOSMA), a separate bill that would set 13 as the minimum age for social media accounts and restrict algorithmic content recommendations for teens. “Chatbot KOSMA,” as he called it, applies the same parent-control philosophy to conversational AI.

What It Means for Windows Families

For the millions of households that run Windows, this isn’t a distant Washington debate. AI chatbots are no longer confined to standalone websites; they are woven into the operating system, the default browser, and the apps children use every day. Microsoft’s Copilot is built into Windows 11 and Edge, offering an AI assistant right on the desktop. Google’s Gemini is integrated into Chrome. Both can be accessed with a few clicks, often without separate account credentials.

If the CHATBOT Act becomes law, platforms will need to institute family account systems that satisfy its requirements. For a Windows-based household, that could mean:

  • The first time your child tries to use Copilot in Edge or a third-party chatbot app, the service must determine they are under 13 (or under 18, depending on the setting) and either block access or route them to a parental consent workflow.
  • Parents would need to manage separate child accounts on those AI platforms, similar to how Microsoft Family Safety already lets adults supervise screen time, app purchases, and web filtering on Windows PCs. The law would extend that oversight to the content of AI conversations.
  • Age verification would become a gateway. The bill text allows for methods like government IDs, financial documents, or other widely accepted practices, raising valid privacy concerns. The House digital-safety framework explicitly says regulations should not force families to hand over government-issued identification, but striking the right balance will be tricky.
  • Enterprise and school devices that run Windows may need to adjust policies if students use AI tools on them. IT administrators will have to evaluate whether existing endpoint management and identity systems can map to the new federal requirements.

Those who rely on Microsoft Family Safety already know the value of separate child accounts for limiting screen time and blocking mature content. The CHATBOT Act would add a new layer: mandatory safety defaults that control how a chatbot can interact with a young user, not just whether the app itself can launch.

How We Got Here: From KOSMA to CHATBOT

The path to the CHATBOT Act is paved with earlier attempts to regulate youth online activity at the federal and state levels.

  • The TAKE IT DOWN Act, co-led by Cruz and signed by President Trump on May 19, 2025, criminalizes the knowing publication of nonconsensual intimate imagery — including AI-generated deepfakes — and requires covered platforms to offer a notice-and-removal process. Its criminal provisions took immediate effect; the platform removal-process requirement is set for one year after enactment.
  • KOSMA (S.278), which Cruz also sponsors, bars social media platforms from allowing anyone under 13 to hold an account, restricts the use of children’s data in recommendation systems, and ties federal E-Rate broadband support to school filtering requirements.
  • In Iowa, state lawmakers debated their own chatbot guardrails earlier this year. One proposal would require chatbots to disclose that they are not human at the start of a conversation and at regular intervals, and to direct a minor toward a suicide hotline when mental-health concerns arise. Iowa Public Radio reported that the legislature struggled with crafting definitions narrow enough not to catch ordinary customer-service bots. That tension is now central to the federal effort.

The CHATBOT Act represents a deliberate evolution: Congress is responding to the unique risks of conversational AI, not simply expanding a social-media template. As Hinson put it, “any place there may be a social interaction” — including gaming platforms like Roblox — could be a venue for risk, and the law must adapt.

What You Can Do Today to Prepare

Legislation is slow; your child may be chatting with an AI today. Here’s how to build a practical safety net on a Windows PC right now.

1. Create separate Windows accounts for children.
Avoid having kids use a parent’s account — it grants unrestricted access to saved credentials, payment methods, and an adult browsing profile. In Windows Settings, go to Accounts > Family & other users and add a child account. This also activates basic activity reporting and screen-time limits.

2. Review browser and app permissions.
Check which AI-powered extensions, sidebars, or desktop apps have access to the microphone, clipboard, file system, or browsing history. In Edge, go to Settings > Cookies and site permissions and audit any AI tools that have broad access. Disable permissions that aren’t essential.

3. Treat AI chats as potentially shareable.
Explain to children that they should not enter addresses, school names, passwords, medical facts, financial details, or photos into any chatbot. Even if a conversation feels private, data can be stored, analyzed, or exposed in a breach.

4. Explain that chatbots can sound confident and still be dangerously wrong.
A fluent answer is not expertise. This is especially important for medical, legal, mental-health, and relationship questions. Set a clear rule: for serious topics — self-harm, threats, sexual content, abuse — the next step is always a trusted adult, not a continued private conversation with an AI.

5. Watch for “secret relationship” dynamics.
Be alert to chatbot products that encourage secrecy, emotional dependency, or the idea that the AI is more trustworthy than family or friends. Some companion-style apps are designed to simulate deep personal relationships and may be marketed specifically to teens.

6. Keep Windows, browsers, and security software up to date.
While patching won’t solve every AI safety problem, it closes known vulnerabilities that can lead to account compromise, malicious extension installs, or data theft. Enable automatic updates and consider using Windows Security’s family options to manage app installations.

For school IT administrators, now is the time to inventory which AI tools students can access on managed Windows devices — including browsers, Microsoft 365 Copilot integrations, and third-party tutoring apps — and to begin planning for age-verification and parental-consent workflows that align with the bill’s likely requirements.

Outlook: When Will These Rules Take Effect?

The CHATBOT Act is still working its way through the Senate committee process; a markup was held in April 2026. Even if it passes both chambers this year, there will be an implementation lag while agencies draft enforcement rules. The TAKE IT DOWN Act’s staggered effective date (immediate for criminal provisions, one year for platform-removal processes) offers a possible model.

What’s clear is that Washington no longer views AI chatbot safety as a niche issue. The Cruz-Hinson roundtable signals that child-protection policy is entering a new phase — one in which the conversational interface of an AI assistant will be regulated with the same seriousness as a social-media news feed. For Windows households, the CHATBOT Act is not the end of the story, but it is the most specific federal blueprint yet for giving parents control over the AI their children talk to.