Hacker News Reader

Comment by apparent | Hacker News Reader

I don't understand how she violated this law:

> Heller faces a charge of making a written threat of violence under Florida law. Florida Statute 836.10 makes it a second-degree felony to send, post, or transmit a written or electronic record threatening to kill or injure someone, carry out a mass shooting, or commit an act of terrorism.

She didn't threaten anything, she wrote down that she was going to do it. A "threat" is more than a mere statement, especially when written in what is described as a "diary".

> A Florida woman is facing felony charges after she used Claude as a diary and allegedly wrote that she planned to "shoot up" the Sheriff's office.

Obviously I don't want anyone to shoot up anything, but this seems like a weak case legally speaking.

View on HN

I think it's fair to say this is a gray area. Clearly it was transmitted.

I can certainly threaten you harm and send it to not-you and you're still clearly in danger even if it wasnt transmitted to you. So the question becomes did she transmit it to someone? Clearly yes she transmitted it to Anthropic. But she clearly intended to send it to Claude, an inanimate object.

I don't think it qualifies for this part:

> The communication must be made in a manner in which another person may view it.

Even 'transmitted' is too broad if you also consider iCloud backup to be a means.

How could it fail to qualify for that? Another person did view it, which clearly establishes that another person could have viewed it.

The clause in the law is pointless, since if you do something that nobody else can see, you can never be punished. But many, many, many laws are written without regard to whether they make any sense.

In general, you have to intend to commit the crime you're being charged with (referred to as "mens rea"). Though, it's important to note that "intend" is extremely ill-defined in the US and it varies with the crime (eg for theft you must take the item on purpose whereas something like manslaughter requires only that you were negligent).

What this means here is, of course, equally spongy, but it is interesting as there might be an argument here that she did not intend for it to be viewed by anyone as, regardless of what the T&C say, most people do not expect their "private" chat logs between them and a machine to be seen by anyone at all.

Too broad. Unless you're transferring ink from a typewriter ribbon onto paper in a hut with no electricity, your words, or my words as I type this, are being grammar checked by something partly in the cloud. If I delete my words, are you saying I've transmitted them nevertheless?

She may have assumed that the chat conversation was private, but it wasn't. She sent a message of intent to harm and a human received the message.

There was no intent for a human to read the message. By your logic, if she wrote a threat in a diary and a burglar broke in and read it, it would be a crime on her part.

Chats with a company's computer aren't private the way a diary is. A better example would be she intended to write it in a word document and instead accidentally sent it in an email to a random person.

Okay so say she wrote it in Office365 in the web browser?

Saying that she “sent a message” is both literally true and obviously intellectually dishonest.

If she had intended to write it in a word document on her computer but instead accidentally wrote it into her email client and sent it to a random person, I'm not convinced she would escape getting charged then either.

It's not any different than telling an automated phone voice tree system that you plan on killing someone and then being surprised that your words were later heard by a human. She absolutely told a company's computer. She sent the message.

The law may have been intended for more direct threats to a person as a means of intimidation, but that's a separate conversation.

Sure, you could intentionally conflate analogies like that if you wanted to mislead.

Yes but the law usually evaluates the application of a statute within the context of someone's mental state. This is why you are not guilty of battery when you trip and accidentally bump into someone. https://en.wikipedia.org/wiki/Mens_rea

That depends on the crime; several related crimes are only distinguished by intent. Negligence is itself a crime if it is the cause of a preventable death when the person has a reasonable obligation, such as when driving a vehicle.

I'm not really sure that this can be likened to a diary when it is called a "chat" but that's for the legal system to determine, not me sitting on my couch.

Any reasonable person presumes when they chat with Claude that it is a computer program on the other end. "Claude is AI" is explicit on the page right under the input box. The word "chat" doesn't anthropomorphize the situation.

And yes, some laws are "strict liability", I don't think this one is.

They didn’t receive it, they secretly extracted it by spying on her.

The Florida statute requires that it be transmitted in a manner that can be viewed by another person. If you have no idea that someone could view your communications with a chatbot, did you really intend to break that specific law? Technically, that threat was communicated to another person but not through her own intentions.

What if she mailed it to herself through the post, and her housemate accidentally opened the mail?

Then they are both innocent of their 'crimes' as they were both unintentional.

She wouldn't have intended her housemate to open it, so it wasn't a threat.

And her housemate didn't intend to open it, so it wasn't Obstruction of Correspondence. https://www.law.cornell.edu/uscode/text/18/1702

A sibling comment includes an important rider to the provision: "...in any manner in which it may be viewed by another person." If you wrote this in a google doc, it almost certainly would not qualify as a threat under this statute. Even though google docs, like LLM chats, have administrative override and you could look at their contents - you would not expect either to be "viewed by another person."

IMO I do not think this is a grey area and it's legal to tell a LLM you want to kill someone. It's certainly not a "threat" like you might send to another person, though it may end up being evidence of conspiracy or premeditation. I suspect we would be well served to, after a few years of experience, put together some laws governing when LLM chats must be made available to authorities.

It is very interesting that the LLM responses to these lines - the context around what she is saying - is not in the article. I suspect, as is the case in many instances where LLMs are involved in violent planning, that the LLM was urging this behavior on. Basically entrapment - you are encouraged by a robot to become more violent and vindictive and then when you do you are handed over to police.

I do take your point. But I would also ask you to imagine that instead of threatening to mass murder a bunch of people she was just creating text based csam. Still, all of what you said applies, but clearly she would be prosecuted over this victimless crime and the jurisprudence in fact disagrees that it would be victimless. So applying that reasoning to this act she clearly committed the crime (legalese notwithstanding). The fact that threatened mass murder doesn't trigger the same thought process is, well at least interesting to me.

I am not sure what point you think you are taking. I have no idea if what I said based on the text of this law applies to the text of a different law. I am having trouble following your argument (especially the "legalese notwithstanding" as my point is based on the legal text and I don't understand the purpose of a conversation around 'is this legal' outside of legal text) - could you say more?

Claude's terms of service makes it very clear that their employees will read messages[0] to determine that they don't contain the things that these messages contained[1].

> Review is needed to enforce our Usage Policy... designated members of our Trust & Safety team may access this data on a need-to-know basis as a part of their evaluation process.

[0]: https://privacy.claude.com/en/articles/10458704-how-does-ant...

[1]: https://www.anthropic.com/legal/aup

I highly doubt she read the terms.

The critical part of a "threat" is that the perpetrator takes some intentional method to deliver it.

I doubt it too, but they're legally binding, and extremely permissive in favor of the company far past the point of technical necessity...

...But since they exist, the company acted more responsibly than if they had simply ignored the data they chose to observe. Which is both the bare minimum given the circumstances, but also strangely absent almost all the time in the industry (including within Anthropic).

I am not a lawyer, but I have the feeling that reporting a discovered terrorist manifesto or something similar is in the spirit of the law.

Nobody is saying that Anthropic didn't have the right to read it. They obviously do.

I am suggesting that the criminal case against her lacks mens rea because inside of her own mind she did not expect that anyone would read it.

A terrorist manifesto is in fact similar -- it wouldn't become a threat until that person takes some action to knowingly communicate it to others -- typically they'd be brought on terrorism/weapons/conspiracy charges.

Doesn't "mens rea" mean "criminal intent"? I can't imagine a better example of this than intent than confessing it through written conversation. Never mind she uploaded it to a service that she explicitly agreed could be checked by employees.

I would love to sue every SaaS company for hard-to-understand terms too, as you suggest, but at what point would Anthropic actually need to tell someone about a terrorist manifesto that their customer gave them to read... Never?

It requires intent specifically to communicate it to someone in a way that is to be interpreted as a threat.

https://www.criminaldefenselawyer.com/crime-penalties/federa...

Whether she intended to act upon the idea might be relevant to some other crime but it doesn't really have anything to do with whether it is a threat.

A reasonable person would not expect humans to review the millions of messages passing through the LLMs, or their own threats to ever be transmitted to a human without their authorization.

Reporting the danger is by itself a good deed. But there should be a better way of restricting firearms from the probably irresponsible lady than using inappropriate charges to punish the thoughtcrime, OR waiting for them to commit violence.

Yeah.. if you write a personal note and it's backed up by the operating system, it appears to be in violation of this law as well (since the company could theoretically read it)

This almost smells like thought crime... Minority Report when?

People assume there won't be another human in the mix, but there is. She was judged for what she probably assumed was a private thought when it was actually not private.

Are thoughts illegal?

Sharing them certainly can be under certain circumstances. I wouldn't be too surprised if this case gets tossed, but it's not inconceivable that prosecutors could win on the grounds that AI chats do not have a reasonable expectation of privacy and are therefor meet the requirements in the Florida statutes, or even more likely, they find some lesser charges for a plea bargain.

Reply on news.ycombinator.com