It is probably not true that ChatGPT has resulted in an increase in murders and suicides, and certainly it would be very difficult to prove liability on OpenAI for this. It reminds me of the campaign in the 90s against video game manufacturers for "corrupting the youth".
But I also don't think they expect to win. They just want to show that they're doing something to fight tech companies and AI.
[1] https://www.cbc.ca/news/canada/british-columbia/sam-altman-t...
There's no legal privacy such as doctor-patient confidentiality or lawyer-client confidentiality. It would not surprise me that people want some sort of guardrails protecting the public from recipes-to-do-evil. Based on how much pushback the AI industry has been doing in response to guardrails about hate speech or porn, I expect this to be fought to the death by the AI industry.
The government did intervene though. They threatened to regulate the industry if the industry didn't regulate itself. So some/all the big industry players got together and created their own independent age rating agency that they all agreed to use.
Whoever was suing won in the outcomes department.
The reputation of Florida comes from having a very broad public records law that requires publication of numerous details of police reports that in other states is kept confidential. That means that sensational stories are much more likely to make it into the news: https://digitalcommons.wcl.american.edu/facsch_lawrev/2117/
My understanding is that OpenAI products specifically provided help in planning attacks / self harm.
The way I see it, providing general information is not a crime. They're basically saying: "Oh no! My repository of all human knowledge contains all human knowledge! It must be defective!"
I’m no lawyer though so maybe potato po-kill your spouse with a claw hammer-tato. They do sound very similar. Please tell me more.
In all the cases I've seen, the user seemed highly motivated to kill themselves and spent a lot of time trying to push past guardrails, ignoring repeated messages to seek help.
Now, AI, data centers, and tech in general are so unpopular that going against them even in a symbolic way is an easy political win on either side of the aisle.
This is the industry that used to have people hyped about iPod and iPhone launch keynotes, lining up at retail stores days ahead of time to experience new technology.
https://www.apa.org/monitor/2026/06/ai-concerns-americans-ad...
Imagine if more than half of Americans thought the iPod mini was bad for society.
I remember when it was 1998 and people in khaki pants were telling us that the information superhighway was going to transport us to a scholastic utopia.
Texas is leaning into becoming the manufacturing and R&D hub for the US, and is courting gigascale data centers and rolling out nuclear power, near-infinite solar, wind, and gas to power it as fast as possible.
Florida is leaning into the retired and populist factions of the GOP, banning data centers and taking on populist anti-tech positions that Texas wouldn't dare (because they want the investment).
Source? It’s been an open secret in academia and medicine that professors [1] and doctors [2] are fleeing Texas’s political climate.
[1] https://www.texastribune.org/2025/09/05/texas-faculty-univer...
[2] https://www.texastribune.org/2024/10/08/Texas-obstetrics-gyn...
Becoming? This has been true for decades in the urban areas
Source? (Not doubting. But I’m finding conflicting figures.)
Florida, at least for local Florida stuff, like what GP is talking about, has had R governor, senate, and house for 25+ years. With a supermajority R for most of that I think.
If so, would this program also open me up to liability in Florida?
const platitudes = ['Good point!', 'You're absolutely right.', 'I agree, let's explore this idea further.', 'This plan is a good idea'];
var prompt;
var response = "Hello, AI here, how can I help you?";
while (true) {
prompt = window.prompt(response);
response = platitudes[Math.floor(Math.random() * platitudes.length)];
}Guns are explicitly exempted from liability rules. They’re the exception that proves the rule.
Yes, but that only eliminates guns as an example of inherently dangerous products which are legally sold without special exemptions. I think the most constructive response is to consider another example without a special exemption - such as nail guns or rat poison.
> They’re the exception that proves the rule.
What rule does guns having a special exemption from (some) product liability laws prove? (serious question, I don't know what you mean.) It doesn't prove dangerous products cannot be sold to the general public without a special exemption. The more useful question is: "since very dangerous products CAN be sold to consumers in some cases, is ChatGPT such a product and is this one of the cases."
Fortunately, there's a highly evolved body of jurisprudence around product liability and negligence to help us tease out these details. Turns out it depends almost entirely on a combination of niggly details like sales and usage context as well as claimed features of the product along with disclaimers, disclosures, existing practice, prior knowledge of actual harm, average user competence, etc. The bottom line is, winning a judgement against OAI in this particular case is probably quite a stretch. But this AG probably doesn't really intend to try this case in court.
I can't sue you for product liability if you strangle me but I can still sue you.
Gun manufacturers have been successfully sued for shootings before [1]; who cares if it's about "product liability"?
[1]: https://hls.harvard.edu/today/a-tough-road-for-suing-gun-mak...
Probably the only response stupider than "Nothing could have prevented this" is "Random thing, other than the mental state of the murderer and the access to firearms, caused this."
Alternative take: The purpose of "thing" is "what it is used for", which is a crude variation of "the purpose of a system is what it does". Reducing it to a single definition is almost always going to be inaccurate.
The way it is used defines it's purpose. The screwdriver was used to open the milo tin so the milo could be removed from the tin. The gun was used to make a hole in the milo tin so the milo could be removed from the tin. Purpose is a per-unique-scenario proposition. The best tool for the job is the one that's available.
To intentionally misquote Arthur Weasley: "What exactly is the purpose of a rubber duck?"
I’ve fired guns. Never to kill things. I’ve also used chat bots to be entirely useless. I wouldn’t endorse this dichotomy of purpose as a basis for any judgement.
I don't think the token providers want the same level of regulation as guns.
The chemtrails conspiracy is just used to dismiss valid concerns about weather modification
Florida doesn't lie? wtf?
Either kids aren't actually being harmed, government regulation will cause more harm, or parents should parent their kids. Either way, nothing about the solution should involve me.
Parents are voters. One of the way they parent is by being civically active in their kids’ interest.
Throw away their TVs and minimize screen time at home[1].
Be responsible for the upbringing of their own children[2].
Learn how to be parents; the government shouldn't force companies to do parenting instead[3].
Not have had children in the first place[4].
Be the ones responsible for parenting their own children[5].
Actually parent their kids and not rely on the government to nanny them[6].
Get to decide what content their children, then like me, you would oppose any kind of legislation with this goal in mind[7].
I could go on. My point is that HN has a long tradition of distrusting regulation especially when it comes to parenting. I have no problem acting as a lightning rod for that arugment.
1. https://news.ycombinator.com/item?id=48182101
2. https://news.ycombinator.com/item?id=48074072
3. https://news.ycombinator.com/item?id=48072708
4. https://news.ycombinator.com/item?id=48069884
5. https://news.ycombinator.com/item?id=47818303
No it isn't. The idea that no amount of actual harm outweighs the possibility of any amount of potential economic harm is not a reasonable one.
Would this lawsuit even be a thing?
That and DeSantis is probably still eyeing a presidential run in 2028 and this will win him some points with his base. This lawsuit is absurd.
Are we reading the same Hacker News comments sections?
aiding and abetting violence: books on the topic since the 5th century BCE
economic disruption: like the printing press
copyright theft: printing tech also makes that far easier
displaces creativity: this was Socrates' objection to reading and writing
misinformation: both techs turbocharge all info, correct or not
environmental impact: e.g. deforestation
amplifies bias: this is a common purpose of writing things down
atrophy of skills: Socrates said reading would damage memory skills
concentration of power: writing was tightly controlled by powerful interests for their leverage and protection
Why? Like, people doing fraud is an instance of the written and spoken word. That doesn’t mean every argument against fraudsters should be leveled against speech.
AI obviously replaces thinking, as can be seen from your comment. No one will refute this point-by-point nonsense.
If it goes anywhere at all, it'll likely just result in a settlement paid to the government and a consent decree mandating well-intended, nice-sounding yet vague rules which just become another compliance cost for leaders, barrier for emerging competitors and otherwise accomplish little of value for citizens. It's also unproductive because it tends to polarize a complex, nuanced and evolving technical issue toward extremes by hijacking it as fodder for existing political and even culture war battles.
While some bad things have certainly happened, proving direct liability under reckless endangerment in court, especially in an area so new, will be virtually impossible. Even willful negligence will be a stretch. This is neither the venue nor instrument of governance we as a society should be using to address these issues. And an attorney general should know that.
Why? There is no incentive to go easy on OpenAI. (Short of Altman stepping down, which he won't.)
> It's all politics, optics
This was my initial reaction. Some excerpts from the complaint [1]. The facts are pretty bad.
Did you follow up on that by looking for any money links between Musk and this AG?
Of course it should steer people away from harmful thoughts like any sensible human would, but that's all you can do, really.
I don't see the state's involvement in that
Can't wait for them to sue the NRA next!
Reverend Doctor Robert Evans had a few episodes on Behind the Bastards this last month about how AI chatbots seem to sometimes create cult-like dynamics with their users. I don't know how this argument will fare in court, but I don't know if this is necessarily wrong.