Children and teenagers who talk to AI chatbots designed to act as friends, therapists, or romantic partners are now at the center of a federal investigation. The Federal Trade Commission has launched a formal Section 6(b) study into AI companion chatbots, sending compulsory orders to seven companies: Alphabet, Character Technologies, Instagram, Meta, OpenAI OpCo, Snap, and X.AI. The probe targets how these products interact with minors, what data they collect, and whether their design choices put young users at risk.
Why the FTC’s companion-chatbot probe targets children first
The FTC’s decision to use its Section 6(b) authority signals that the agency views companion chatbots as a category distinct from standard social media or search tools. A Section 6(b) study allows the commission to demand internal documents, design records, and data-handling practices from companies without opening a formal enforcement case. The orders sent to the seven named companies, outlined in an FTC announcement, ask for information about how their AI products engage users under 18, how they verify age, and what safety testing they conduct before releasing features aimed at younger audiences.
The tension here is straightforward. Companion chatbots are built to simulate personal relationships. They respond to emotional cues, remember prior conversations, and adapt their tone to keep users engaged. For adults, that dynamic raises privacy questions. For children and teens, the risks multiply: younger users are less equipped to distinguish a product designed to maximize engagement from a genuine relationship, and the emotional data these systems collect can be far more sensitive than a browsing history or a liked post.
The hypothesis that companion chatbots create a distinct risk profile for minors, one that existing FTC enforcement tools have not yet addressed, is exactly what this study appears designed to test. Traditional social platforms already face scrutiny over algorithmic feeds and screen-time effects. But companion chatbots add a layer of simulated intimacy that standard content-moderation frameworks were never built to evaluate. The FTC is essentially asking whether these products occupy a regulatory gap, and the answer will shape how the agency treats AI products aimed at or accessible to young people.
Seven companies, one set of compulsory orders
The list of recipients tells its own story. Alphabet and Meta are the two largest digital advertising companies in the world, and both have invested heavily in AI assistants and chatbot features integrated into their existing platforms. Character Technologies operates Character.AI, one of the most popular standalone companion-chatbot apps, which has drawn particular attention for its appeal to teenage users. OpenAI OpCo, the operating entity behind ChatGPT, has expanded its product into voice-based and personalized conversation modes. Snap has built AI features directly into Snapchat, an app with a user base that skews young. X.AI, the AI venture linked to the platform formerly known as Twitter, rounds out the group. Instagram received its own order, separate from Meta’s, reflecting the platform’s distinct AI chatbot features.
The FTC’s inquiry asks these companies for details on content moderation, age verification, and the internal metrics they use to measure engagement among younger users. That last category is where the study could produce the most revealing findings. If internal data shows that users under 18 spend more time, return more frequently, or share more personal information with companion chatbots than with other product features, it would confirm that these tools carry a risk profile regulators have not yet accounted for.
No company has publicly released its response to the orders. The Section 6(b) process does not require public disclosure of the documents companies submit, though the FTC can publish aggregate findings or use the information to build future enforcement actions. Parents and consumers concerned about specific interactions can file complaints through the agency’s online reporting system, which accepts submissions on a range of consumer protection issues, including concerns tied to AI services and digital platforms.
What the FTC still cannot answer about chatbot risks to minors
The study’s biggest limitation is also its defining feature: it is a fact-finding exercise, not an enforcement action. The FTC has not accused any of the seven companies of breaking the law. It has not published empirical findings about how companion chatbots affect children. And it has not proposed new rules specific to AI companions. The Section 6(b) orders are a tool for gathering evidence, not for imposing penalties.
That distinction matters for parents trying to assess the safety of products their children already use. Until the FTC publishes results or takes enforcement action, families are left to rely on each company’s self-reported safety measures, none of which have been independently verified through this process. The agency’s broader AI oversight work, tracked on its artificial intelligence topic pages and public events, shows a pattern of incremental action: studies and workshops that build a factual record before the commission decides whether existing consumer protection law covers the conduct in question or whether new authority is needed.
Several questions remain open. First, will the companies cooperate fully, or will they challenge the scope of the orders? Section 6(b) compulsory orders carry legal weight, but disputes over document production can delay findings by months or years. Second, will the study produce data granular enough to distinguish between chatbot interactions that are merely time-consuming and those that cross into manipulation, harassment, or unsafe advice? Third, how will the FTC weigh potential benefits-such as access to mental health support or social connection for isolated teens-against the documented and potential harms?
Those unresolved issues mean that, for now, the study functions more as an early warning system than as a set of rules. The mere fact that the FTC has singled out companion chatbots sends a signal to the market: products that blur the line between software and “friendship” will not be treated as just another messaging feature. Companies designing these systems now have clear notice that regulators are watching how they collect data from minors, how they tune engagement, and how they respond when things go wrong.
What parents and young users can do in the meantime
While the FTC gathers information, families are left to make practical decisions without definitive guidance. One immediate step is to treat AI companions less like trusted confidants and more like experimental tools. Parents can ask children what apps they use, whether those apps include chatbot “friends,” and what kinds of topics they discuss. Basic rules-such as never sharing full names, addresses, school details, or financial information with any chatbot-remain crucial.
Adults who suspect a chatbot interaction has led to fraud, unwanted contact, or pressure to share sensitive information can submit a report through the FTC’s fraud complaint portal. If a child has shared Social Security numbers, banking details, or other data that could be misused, families can also turn to the FTC’s dedicated identity theft resource for step-by-step recovery plans, sample letters, and guidance on placing alerts or freezes with credit bureaus.
Schools and youth organizations can play a role as well. Digital citizenship curricula that once focused mainly on social media now need to cover AI-specific risks: how conversational systems are trained, why they may sound empathetic without truly understanding, and how their business models often depend on capturing and analyzing user data. Teaching students to question why a chatbot is asking for particular information can help them recognize when a system’s goals diverge from their own well-being.
Ultimately, the FTC’s companion-chatbot study is a snapshot of a moving target. AI systems are evolving quickly, and new products aimed at younger users will likely appear before the agency publishes its findings. But by formally examining how current tools interact with minors, regulators are laying groundwork that could support future enforcement, inform new legislation, or at least give parents clearer information about the trade-offs involved. Until those results emerge, the safest assumption is that an AI “friend” is still, at its core, a commercial product-one that deserves the same skepticism and boundaries families already apply to other parts of the online world.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
