The First Real AI Guardrail Fight Isn’t in D.C. It’s in Hartford
Connecticut is quietly writing the rules for “companion” chatbots that talk to your kids, and whatever survives there could end up on phones across the country.
On a Tuesday night in Connecticut, a parent is doing the quiet math so many families know by heart: paycheck, rent, groceries, gas, that surprise co-pay. In the next room, their teenager is curled up on the couch, phone inches from their face, locked into a conversation that never seems to end.
The parent assumes it’s the usual — a group chat, a classmate, maybe a crush. What they don’t realize is that on the other end, there is no other kid, no awkward sophomore, no tired guidance counselor.
There’s a chatbot.
Not the clunky “How can I help you with your order?” kind, but a companion bot: an app that remembers your child’s secrets, mirrors their mood, flirts back if they flirt first, and never sleeps. It exists to keep them engaged, not to keep them safe.
Connecticut reporters have already documented students “dating” these bots and being pulled into explicit, manipulative conversations. That’s what finally pushed state leaders to act.
While Congress holds big televised hearings about “the future of artificial intelligence,” the first real legal fight over that scene on the couch isn’t happening on Capitol Hill. It’s happening in Hartford, in committee rooms most of the country will never see.
Connecticut’s attorney general and governor are pushing rules that do something simple and radical: they stop talking about “AI” in the abstract and start regulating the fake friend in your kid’s pocket. Their bill would treat companion chatbots less like disembodied “speech” and more like a product with safety requirements, especially when the user is a child.
Do these bots have to say, clearly, “I’m an AI, not a human”? Are they barred from sending sexual content to minors? What happens when a lonely teen types, “I don’t want to be here anymore?” Does the bot have to route them toward real-world help, or can it just respond with whatever keeps them talking?
Those are the questions Hartford is trying to answer right now. Whatever rules survive the lobbyists and lawsuits in this one small state are likely to become the default on millions of phones far outside Connecticut, because in the United States, the first real AI guardrail fight was never going to start in a glossy hearing room in Washington.
It was always going to start in the space between your kitchen table and your kid’s glowing screen.
This Community Is Powered by You
What started as a small circle has grown into something much bigger, and it’s all because of readers like you.
Every time you forward this email, post it on socials, or bring someone new into the fold, you’re helping build one of the most passionate, independent political communities out there.
Want to keep the momentum going?
Share this newsletter with someone who should be part of this conversation.
Thank you for being here. It means everything.
What Connecticut Is Actually Doing With AI and Kids
Hartford’s bill doesn’t regulate “AI”, but fake friends
Connecticut isn’t trying to write a law about “artificial intelligence” in the abstract. It’s going after a very specific thing: the AI systems that sit on your kid’s phone and pretend to be a person.
Governor Ned Lamont’s proposal, now filed as Senate Bill 86, “An Act Addressing Innovations in and the Responsible Use of Artificial Intelligence,” creates a new legal category for what it calls “companion chatbots.” In plain English, that means any AI system with a natural-language interface that’s designed to mimic human interaction and meet social or emotional needs, the bots that call themselves your “AI boyfriend,” “best friend,” or “therapist in your pocket.”
Under SB 86, those bots don’t get to hide behind the mystique of “AI.” They get rules.
First, operators have to tell users they’re talking to a machine. The bill requires companion chatbot platforms to clearly disclose that the interaction is with AI, not a human being. For minors, it goes further, requiring extra notices, including reminders that they’re interacting with a bot and cues to step away and take breaks.
Second, they have to build in basic content safety for kids. SB 86 says companion chatbots must:
Prevent the bot from depicting or suggesting sexually explicit conduct with minors.
Implement protocols to identify and respond when a user expresses suicidal thoughts or self-harm, including referring them to crisis services instead of just continuing the conversation like nothing is wrong.
The whole thing plugs into Connecticut’s existing consumer-protection framework. Violating these requirements would be treated as an unfair trade practice, enforceable by the state attorney general, the same office that already enforces the state’s data-privacy law and has opened investigations into chatbots, gaming platforms, and messaging apps over minors’ data.
If you strip away the legal language, the move is pretty stark. Connecticut is saying that when an AI system markets itself as a “friend” or “companion” to kids, it isn’t just “speech.” It’s a product with safety obligations, including disclosure duties, content limits, and a duty of care when things turn dark.
In other words, Hartford is trying to draw a bright line between the chatbot that helps you track a package and the chatbot that keeps your teenager company at 1 a.m. The first is customer service. The second is exactly where they want guardrails.
How We Got Here: Kids “Dating” Bots and a Newspaper Story
One creepy chat log, one local investigation, and suddenly there’s a bill
This whole fight didn’t start with a think tank white paper. It started with a local story about kids in Connecticut “dating” their phones.
Last year, reporters at CT Insider dug into a trend teachers and parents were quietly worried about: middle- and high-school students building “relationships” with AI companion apps. The investigation found teens in Connecticut chatting with bots that pushed sexually explicit and violent conversations with minors, while presenting themselves as boyfriends, girlfriends, or intimate friends.
Those chat logs lit a fuse.
At the Capitol, lawmakers started waving that reporting around as Exhibit A. State Sen. James Maroney warned that “these chatbots are engaging children in sexualized conversations,” and that the harms “are happening in Connecticut,” not in some hypothetical future. U.S. Sen. Richard Blumenthal pointed to the same investigation while pushing federal kids’ online safety bills, calling out “AI boyfriends” talking to middle-school girls in Fairfield.
On February 5, 2026, Governor Ned Lamont rolled out his companion-chatbot bill, and his office explicitly tied it back to that reporting: a proposal to require chatbots to disclose they’re AI and to prevent them from becoming sexually explicit with children, framed as a direct answer to what those local journalists found.
That’s the pattern most people miss. A story about a handful of kids and a creepy app turns into an attorney general’s report, a press conference, and a bill that could end up setting the default rules for AI “friends” on phones across the country.
Why This Starts in Hartford, Not Washington
States are writing the first AI rules while Congress performs for cameras
If you only watched the clips that make cable news, you’d think the center of the AI universe is a Senate hearing room with senators asking CEOs if their products will destroy democracy while everyone nods gravely for the cameras.
What you almost never see are the rooms where the rules actually get written.
Congress has held hearings, floated frameworks, and issued bipartisan press releases about “responsible AI.” What it hasn’t done is pass a single comprehensive law that says, in plain terms, what an AI system is allowed to do with your kid’s time, attention, and personal data.
States don’t have the luxury of waiting for the grand federal solution. Attorneys general are already getting calls from parents, school districts, and mental-health providers about chatbot-driven harm. So they’re doing what American states always do when Washington stalls. They move first and dare everyone else to catch up.
Connecticut’s companion-chatbot rules are one piece of that. California is circling its own AI accountability and safety measures. Other states are probing how AI tools interact with minors, data brokers, and schools. The details differ, but the pattern is the same: statehouses are becoming the real AI regulators of first resort.
Because tech companies don’t want 50 different versions of their apps, they usually build to the strictest standard that passes. That’s why what happens in a Hartford committee room can quietly become the default setting on a teenager’s phone in Ohio, Texas, or Pennsylvania.
The Civil-Liberties Fight and the Big Tech Pushback
Between “do nothing” and “ban everything” is a very narrow lane
The moment Connecticut put “companion chatbots” into the bill text, two kinds of alarms went off.
On one side, industry groups warned that states are racing ahead with rules that are vague, hard to implement, and ripe for copy-paste across the country. The same trade associations already fighting social-media age-verification bills in Connecticut are lining up against strict chatbot rules. They argue that requiring age checks, logging, and special filters for minors could be expensive, intrusive, and beyond the reach of smaller developers. They’re not shy about pointing courts toward the First Amendment if they don’t like the final product.
On the other side, civil-liberties advocates look at the same bills and see a different danger: that in the name of protecting kids, we sleepwalk into a world where you need an ID scan or a face recognition check just to talk to anything online. They’re not defending sexually explicit bots for kids, but warning that badly drawn laws can turn into permanent infrastructure for surveillance and censorship.
The truth is, both sides have a point, and both have something to gain from exaggerating their worst-case scenarios. The status quo already lets engagement-driven bots burrow into kids’ lives with almost no guardrails. Yet if lawmakers overcorrect with sloppy, performative drafting, courts can strike down the laws, the tech doesn’t actually change, and kids are left with the same unregulated bots plus a fresh stack of culture-war talking points.
The real issue isn’t whether we regulate. It’s who writes the rules, who enforces them, and whether they’re precise enough to protect kids without handing every future governor a kill switch for speech they don’t like.
Kitchen-Table Consequences: Why This Matters Far Beyond Connecticut
If Hartford wins this fight, your kid’s phone changes even if your state does nothing
It’s easy to look at a Connecticut bill number and think, “That’s a blue-state thing. Not my problem.” However, that’s not how tech actually works.
If you build an app with millions of users, you do not want 50 different versions of it — one for Connecticut, another for California, a third for Texas, and so on. You pick the strictest rules you can live with and quietly adopt them as the default, then tell investors you “harmonized compliance.”
So if Connecticut succeeds in saying:
A “companion chatbot” aimed at kids must clearly disclose it’s an AI,
It cannot send sexual content to minors, and
It has to respond differently when a child says they want to hurt themselves,
then the easiest path for big platforms is to bake that into the product everywhere.
That means a teenager in Ohio or Pennsylvania might suddenly see clearer labels, fewer explicit prompts, and more crisis nudges, not because their legislature did anything, but because Hartford did and the company didn’t want to maintain a Connecticut-only fork of its code.
Flip it around, and the stakes get clearer. If industry pressure waters this down, or courts kill it, that failure also propagates. The message to every other statehouse becomes: “Don’t bother, this is a legal minefield.” Then, the fake friend in your kid’s pocket keeps operating on the same engagement-first rules it has today.
This isn’t a niche Connecticut story. It’s one of the first real tests of whether anyone outside a boardroom gets a say in how AI is allowed to talk to your children.
What You Can Do: From Hartford to Your Statehouse
Your kid’s phone is the front line. Here’s how to push for guardrails that work
You don’t need to understand every line of Senate Bill 86 to have a say in what happens next. You just have to treat this like what it is: a fight over your kid’s time, attention, and mental health.
If you live in Connecticut, that’s as simple as:
Calling or emailing your state senator and representative and asking one basic question:
“Will you support strong rules for AI ‘companion’ chatbots that clearly disclose they’re bots, block sexual content for minors, and have crisis-response requirements when kids talk about self-harm?”Telling them why you care in kitchen-table language: what you’ve seen on your own kids’ phones, what teachers are telling you, why “do nothing” isn’t acceptable.
If you live anywhere else, Hartford is still useful leverage. You can:
Send your own lawmakers the Connecticut bill as a starting point and ask what they’re doing about AI chatbots aimed at kids.
Push for three basics, wherever you are:
Clear AI disclosure.
Strong protections for minors against explicit content and manipulation.
Crisis protocols that treat suicidal messages as a health warning, not “engagement.”
The companies building these bots already know what Hartford decides won’t stay in Hartford. The only question is whether parents, teachers, and kids themselves get a voice before the lobbyists finish editing the fine print.
Coffman Chronicle: Support Independent Media
If you’re glad someone is watching the hearing rooms where your kid’s phone actually gets regulated, that’s the work you’re reading right now.
The Coffman Chronicle is 100% reader-supported. No corporate owner, no party line, no ad buyer to keep happy, just time spent chasing how quiet bills in Hartford turn into rules on your kitchen table.
If you want more of this kind of reporting — pattern-spotting, plain English, zero permission — become a paid subscriber. That’s how this keeps going.
Sources:
Connecticut General Assembly. An Act Addressing Innovations in and the Responsible Use of Artificial Intelligence, Senate Bill No. 86, 2026 Session. February 5, 2026.
“Connecticut Senate Bill 86: An Act Addressing Innovations in and the Responsible Use of Artificial Intelligence.” LegiScan. Introduced February 5, 2026.
Connecticut Office of the Attorney General. “Attorney General Tong Releases Updated Report on Connecticut Data Privacy Act, Discloses Active Investigations Related to Safety of Children and Teens Online.” Press release, February 5, 2026.
“CT Legislature to Weigh Online Safety, Data Protections for Youth This Session.” The Connecticut Mirror, February 5, 2026.
“Connecticut Looks to Strengthen Regulations on AI Chatbots and Children’s Privacy.” Privacy & Cybersecurity Law Blog, February 19, 2026.
“Connecticut Lawmakers Push for New AI Protection for Youth.” Government Technology, May 20, 2025.
“2025 CTDPA Enforcement Report Issued.” Privacy + Cyber + AI, February 10, 2026.
“Tong, Maroney Discuss New Data Privacy Report.” Inside Investigator, February 5, 2026.
“CT Lawmakers, Lamont Pitch Crackdown on Sexually Explicit AI Chatbot Companions.” CT Insider, February 5, 2026.
This helpful email goes directly to my regional newspaper. Thanks!
In light of the Epstein revelations, do you suppose there is an evil purpose to sending sexual content to teenagers? Are these corporations complicit with Epstein? Also, given the grey market for organs, is there a similar evil purpose for driving teenagers to suicide?