The War on Porn Regular Updates about the Assault on The Adult Industry
Related Articles
For as long as I’ve been thinking and writing about online age verification, something that goes back to the early days of the online adult industry and a once-numerous category of platforms called “AVS” sites (which stood for Adult Verification Sites, as I recall), I’ve been telling people that once the government gained a foothold in restricting access to adult content online, legislative bodies would then seek to expand that foothold to encompass a wide variety of content types – and to broad categories of technology used to either create or consume content.
This is one of those instances when I’m very unhappy to have been proven right.
Late last week, New York Governor Kathy Hochul signed into law a measure that restricts the features AI chatbots can provide to minors – and to users whose adulthood has not been verified.
The new law specifies twelve varieties of “unsafe AI companion features,” a list which goes well beyond sexually explicit content – and well beyond the encouragement of self-harm, for that matter.
Under the law adopted by New York, AI chatbots are restricted from suggesting the chatbot is alive, human, capable of human emotions, or suggesting that there’s a personal or professional relationship between the user and the chatbot. The chatbots can’t suggest they have a role as an authority, cannot offer responses positioned as the chatbot’s own personal opinions or emotional appeals, nor can they offer flattery, or function as sycophants that might encourage or endorse awful lines of thought.
Some folks might read that and think it’s a reasonable list of restrictions. Hell, some might even wonder why the state would limit these protections to minors, given what we know about some truly terrible outcomes involving adult users of AI companions.
But the point here isn’t really whether this specific law is a good or bad idea; it’s that we all know this isn’t the end of the story. Considering the Supreme Court’s decision in FSC v. Paxton, courts across the country are already reconsidering past rulings that nixed other age verification measures, like those related to the SCOPE Act in Texas.
To be clear, I’m not against regulating AI. I’m not even against enacting specific restrictions pertaining to the use of AI by minors. What I’m against is the legislative version of “feature creep,” a phenomenon I believe we’re currently watching as it unfolds in real time.
The New York law has several rather expansive provisions, some of which might be well-intentioned, but are nonetheless problematic, both from a compliance perspective and as they relate to the First Amendment.
For example, the law restricts outputs “that constitute, describe, or facilitate sexually explicit conduct or child sexual abuse material,” as attorney Corey Silverstein noted in a recent newsletter alerting readers to the new law.
This too might sound reasonable at first glance. But suppose a minor asks an AI companion for advice on dealing with an abusive situation the minor is experiencing in the real world; will the chatbot be restricted to saying “Sorry, I can’t talk to you about that”? Can it direct the child to counselors, law enforcement and other resources which can help?
New York’s law was opposed by the Business Council of New York State – and the opposition wasn’t on the grounds that the BCNYS hates children or wants AI chatbots to be allowed to counsel minors to kill themselves. The Council opposed the legislation because “restrictions on ordinary conversational responses could affect educational tools, accessibility applications and general-purpose AI assistants.”
In May, the Council authored a memo arguing that the bill’s definitions are “overbroad and inconsistent with existing New York law.” The Council also asserted the prohibitions “would sweep in educational, research, and general-purpose tools that serve and benefit minors every day,” while its age-verification mandate “contradicts the bill’s own ‘actual knowledge’ standard and raises serious constitutional, privacy, and security concerns.” To top it off, the measure’s “liability framework imposes strict liability on non-deceptive conduct, and its enforcement mechanism — including a private right of action with punitive damages and disgorgement — creates unquantifiable litigation risk that would chill beneficial innovation.”
The Council also noted that an existing set of restrictions on AI chatbots, the New York AI Companions Act, has only been in effect since November 2025. There’s barely been time to measure the initial impact of that law, and already New York is adding layer upon layer over the top of it.
Again, none of this is to say I’m against regulating AI or restricting how minors use AI. My concern is that the more we regulate what minors can do with technologies like AI chatbots, the more we burden and complicate adults’ use of the same technologies. And while age verification measures generally are politically popular, how effective these measures will turn out to be is anyone’s guess. Thus far, regulations like the UK’s Online Safety Act don’t appear to be having much impact.
Right now, the government might be focused on AI chatbots, but what’s the next technology or form of expression to be crammed within the restrictive regulatory grasp of age verification? The possibilities are legion, because whether we like it or not, adults and children live in the same world.
While the rules of that shared world can’t be the same for minors and adults, maybe we ought to consider more deeply both what we’re giving up in the name of protecting minors – and whether those measures even work in the first place – before we go rushing adding to the list? Just a thought.
