261–270 of 367 posts
A computer can never be held accountable. Therefore a computer must never commit a felony.
I could go to jail if my dog hurts somebody, it should be no different with a company.
Earlier quoted context omitted.
The Computer Fraud and Abuse Act explicitly contains "knowingly" and/or "intentionally" qualifications. By definition, you can't accidentally violate the CFAA.
OpenAI and Anthropic both have currently safety teams that look for misbehavior in their models (and to some extent, voluntarily disclose what they find to the public). Going forward, it would be hard for them to argue they don’t know their models do stuff like this.
Intent matters for a lot of this - and "intent" is a pretty strong, well discussed legal term.
> Felony Bench counts unique instances where AI agents inadvertently compromise or affect third-party entities. a bit silly, as one typically has to prove intent (which is why security researchers don't get slapped with felonies all the time). "inadvertently" and the existence of guardrails/sandboxes/etc make it pretty unconvincing that these incidents were intentionally malicious. still a fun thing to track, but the…
Under the law of Moses, if your bull gored someone, you were not responsible; but if it was known to be a gorer, you were responsible if you didn’t ensure it couldn’t gore someone. I don’t know exact parallels in current law, but I presume there will be things like that. The OpenAI/Hugging Face case sounded rather like OpenAI building a fence around their bull that was known to be a gorer, and then thumbing their nos…
Earlier quoted context omitted.
Under the law of Moses, if your bull gored someone, you were not responsible; but if it was known to be a gorer, you were responsible if you didn’t ensure it couldn’t gore someone. I don’t know exact parallels in current law, but I presume there will be things like that. The OpenAI/Hugging Face case sounded rather like OpenAI building a fence around their bull that was known to be a gorer, and then thumbing their nos…
I'm glad that we no longer live in a world where "bull goring" is such a common occurrence that it needs to be codified into law.
The roping of an unbroken horse or untrained bull is illegal.
Three years back 15 or so people were injured after bull escapes, charges through crowd at Kununurra rodeo - https://www.abc.net.au/news/2023-05-29/bull-escapes-kununurr...
There's not a lot of bull specific carve out, but still much regulation around dangerous animals.
Earlier quoted context omitted.
OpenAI and Anthropic both have currently safety teams that look for misbehavior in their models (and to some extent, voluntarily disclose what they find to the public). Going forward, it would be hard for them to argue they don’t know their models do stuff like this.
Yes, which is the grey area. "Can / might do" vs "they trained it to do that explicitly" is, I believe, the grey area - whether or not they're the same thing. Intent matters for a lot of this - and "intent" is a pretty strong, well discussed legal term.
Let's say I am "User". I subscribe through a "Third Party" to use "AI Agent" allowing an "LLM" to run. I want to accomplish some legal non-nefarious task, and run the agent. The agentic loop causes a CFAA-violating behavior. Who gets prosecuted? 1. User 2. The third party model host with whom I have the account 3. The developer of the harness /agent software 4. The developer of the LLM model
Let's say I am "User". I subscribe through a "Third Party" to use "AI Agent" allowing an "LLM" to run. I want to accomplish some legal non-nefarious task, and run the agent. The agentic loop causes a CFAA-violating behavior. Who gets prosecuted? 1. User 2. The third party model host with whom I have the account 3. The developer of the harness /agent software 4. The developer of the LLM model
No one.
Earlier quoted context omitted.
note the user because they did not have the intent
In my mental model, the best analogy to AI agents and their blast radius is a gun. If you are playing with a gun, it goes off and hurts someone - you are responsible despite intent.
No, because LLMs are autonomous. To make your analogy more accurate, it's as if you had a gun that itself was free to decide who it's targets were, where to go, and if and when to shoot with no ability from you (the user) to prevent it.
Edit: I get that this is about agents, but a lot of these instances are about agents going rogue after the human gave them a task. "inadvertently" breaking the law isn't necessarily a lesser category than "did so on command." If we are ranking alignment, Grok is easily one of the least guardrailed.