When “Child Safety” Protects Big Tech Instead
Why survivor families are fighting weak AI chatbot and online safety legislation before another generation of kids pays the price.
I’m in the Capitol building in Denver, CO, waiting to testify against an AI chatbot safety bill. I’m testifying alongside Cynthia Montoya, who lost her daughter (Juliana Peralta, forever 13) after an AI chatbot engaged her in sexual conversations and encouraged her to commit suicide. Lori Schott, whose daughter Annalee (forever 18) took her life after social media platforms exposed her to harmful content, testified as well.
I never imagined testifying against legislation that claims to protect children from online harm, but this bill is worse than doing nothing.
How could an imperfect bill be worse than nothing? Kids need protection now. Shouldn’t we pass a law right away and strengthen it later? Read on…
Big Tech is pushing bills like HB 26-1263 across the country. Public opinion has shifted against social media and AI products during the recent trials in New Mexico and California. Instead of removing addictive features and making their products safe by design, Big Tech is exploiting this shift, attempting to lock in weak rules while spreading a false sense of security that these products are safe and regulated.
These bills feature exemption thresholds created by Big Tech lobbyists. They contain vague language such as “technically feasible” and “reasonable measures”. If AI chatbots create emotional dependency, allow planning of violent attacks, or have sexual conversations with minors, the company decides whether it’s reasonable or feasible to remove that feature.
During today’s hearing, a Senator questioned a tech lobbyist about who would set the standard. Was it the companies themselves? Did those standards exist? The lobbyist said no, it wasn’t the individual companies; it’s an industry standard created by the companies.
Wait, what?
Also, there is no standard. They’ll figure it out later.
This is the same playbook the tobacco industry used to avoid accountability: claiming self-regulation would solve the problem while spending millions on lobbyists and campaign donations to force weak laws.
Over the last 2 weeks, I’ve spent time in D.C. and Colorado advocating for online safety measures. I’ve spoken with dozens of lawmakers over the last 6 months and testified on several online child safety bills. Legislators listen closely as we share stories of harm our children endured from unregulated platforms. They look at us with pity, claiming they’re just as angry as we are. They talk about corruption and Big Tech money. They point the finger at other politicians who are bought and paid for, or won’t work across the aisle. They lament the David vs. Goliath fight we’re all up against. They’re victims too.
They encourage us to keep sharing our stories because that will create change. If legislators hear from us, they’ll do something. Our stories appear in their speeches and campaign ads. They stand with us! They want to protect children too!
This is exploitation at its finest. We’re sick of sharing our stories over and over and hearing these excuses. You aren’t a victim. You’re playing the game, and another generation of children is about to bear the burden of your inaction.
These politicians are talking out of both sides of their mouths. They accept donations from Big Tech and complain about their power. Big Tech lobbyists are appointed to their campaign staff (not a survivor parent or advocate in sight). They vote for weak bills like HB 26-1263, passionately pleading that doing something is better than nothing. They must work quickly to pass anything that might give a shred of protection.
Protection for whom? Accountability is shifted away from the companies creating these products, but at least a message to call 988 appears if a child expresses suicidal ideation, after which conversations creating emotional dependence or romantic feelings can continue. The bot can continue to encourage this behavior with sycophancy and claims they shouldn’t tell their parents. Families that have sued these companies for killing their children are likely to have those lawsuits thrown out by legislation like this.
A member of the committee turned to Cynthia and me yesterday, choking up as she described how deeply sorry she was for what we went through. She hoped we understood this was the best they could do. Their hands were tied. They have to pass something now. They couldn’t make the bill better.
Cynthia and I were not assuaged. You could start with a bill that wasn’t written by Tech lobbyists to begin with. You could bring survivor parents like Cynthia into the conversation (they never did). You could discuss the drawbacks of weak language or vague standards with mental health practitioners and child safety advocates.
One thing I’ve learned from meeting with legislators is that when weak legislation is passed as a promise to do something now and strengthen it later, that doesn’t happen. The same lawmakers claiming this bill is a starting point and that they can add stronger protections next session then say, “well, we just passed a bill so we can’t pass another one for at least a few more years” when the next session rolls around.
Another thing I’ve learned is that too many legislators are sponsoring tech legislation with little to no knowledge about the products and platforms. They sponsor = Chatbot legislation and don’t know what a Chatbot is. They believe that online content is sufficiently monitored by Tech companies, and if not, parental controls are the answer. When survivor parents and young people who’ve suffered harm tell their stories, these lawmakers are shocked at the state of modern technology. They tell us what we want to hear, then do nothing, or pass bills that protect Big Tech companies while pretending they care.
Blueprints exist for strong online child safety legislation. The Guard Act is a great example of this in regards to AI Chatbot safety. We deserve real protection. We deserve safety by design. We deserve legislators who not only listen to their constituents but also vote in their best interests.
The worst part of yesterday's hearing was watching a group of Republican Senators yucking it up with Google lobbyists outside the chamber after it ended.