No, because clerks are capable of learning from their mistakes. LLMs aren’t capable of knowing they made one, much less learning from it.
You are trying to equate people with things.
Yes AI is a thing, a thing I use to find relevant cases and statutes. Which I then read and review.
Just like 30 years ago I used things called research books to find relevant cases and statutes , which I then read and reviewed.
Those research books no longer exist. They’ve been replaced by online finding sources that use AI, in various forms.
Guess I can no longer practise law.
If you use “AI” then you aren’t, you are just fooling around with lies and hallucinations.
Moderating:
You have made your point. You are now attacking posters. You also seem to be ignoring everything other posters have said in this thread.
Drop it now. If that’s all you have to say, you can leave the thread.
What difference does this make? To a client or the court if a mistake goes through uncorrected - leading to potential negative consequences - why do they care if it was made by a clerk whose work wasn’t checked or AI whose work wasn’t checked?
I attended a seminar recently at which both a judge and a major client (an insurer) stated they not only accepted but expected lawyers to use AI. In the former case they said they expected it to happen but were OK with it as long as the lawyers signing off on the work understood they needed to check AI output and would be held responsible if it was wrong. In the latter case they explicity stated they understood that using AI for (say) discovery or summary of documents may lead to errors but that their expectation was that this would permit some errors to go through but that overall they would effect savings in legal costs.
Your blanket, un-nuanced position that lawyers using AI should be punished is completely out of step with every major stakeholder.
I don’t even like AI and never use it. I know reality when it’s slapping me in the face, though.
Moderating:
It’s really not fair to ask a poster to reply after he’s been told to drop a topic. Please don’t do that.
This is what punishment for professional abuse of AI should look like.
No more trivial fines. “You wasted my time, you assholes, so now I’ma blow up your careers.”
The whole memo is worth reading. 20 pages, allow 10 minutes. It’s a doozy.
Thanks for this. I’ll be circulating it to some students.
She admitted that she used “Grok”, an external AI tool, to assist in drafting and research without verifying the accuracy of the output.
I mean, that right there ought to be a mandatory disbarment. Using Grok for any purpose whatsoever, I mean.
On an even more serious note, the firm apparently had an AI policy that she was aware of and affirmed the would abide by, but did not (had she, she would have been limited to one specific AI tool that was, presumably, associated with a legal database—and obviously checked the work put out regardless).
And although she says the right words to appear contrite in front of the court, the firm had in fact been notified previously (by opposing counsel) that she had made similar gross errors in at least one other case. Clearly, that was not enough. Myself, I’d be mortified if I misrepresented a single citation in a single brief before a single tribunal. Let alone multiple citations in multiple briefs before multiple courts over the course of months or years.
Probably why they came down hardest on her (Norris and Watson were effectively her supervisors, and still had an ethical obligation to ensure he work complied with rules of conduct for attorneys).
Having read the decision all the way through, I don’t disagree with any of it.
I also notice this point that the Court here quotes, which summarises my views on AI:
AI is a powerful tool, that when used prudently, provides immense benefits. When used carelessly, it produces frustratingly realistic legal fiction that takes inordinately longer to respond to than create.
I also like what the judge identifies at the end as the reason for taking this matter so seriously:
At the end of the day, it is not the Court who is the victim but the people whose day in court is endlessly delayed. This cannot continue.
In this particular case, a jury trial had to be cancelled and all other proceedings in the case have been on hold while this matter got straightened out. That potentially causes prejudice to both the plaintiff and the defendant while they wait for their case to proceed.
I like the sentence that introduces the analysis section, just before that quoted excerpt.
Lay people who have only casual experience with the field of law tend to think of motions and decisions and such as being jargon-heavy and dryly impenetrable, but some professionals are actually quite skilled writers who can produce entertaining and accessible documents despite the technical language requirements. This judge is clearly one of the good writers.
I’m not sure if it is because my practice has given me such a poor view of so many lawyers, but I’m not as outraged by this as many of you. I regularly see such incompetent and IMO unethical efforts by attorneys, with such little effective disciplinary action, that reliance on AI hallucinations does not strike me as significantly worse.
I suppose my area of practice is different than some, in that specific case citations are rarely determinative. Needn’t get into a lengthy discussion of the value of precedent, but everyone knows what the major cases say. If someone cites something wacky, it is easy to check. No lawyer or judge would just accept it. And out-of-circuit precedent is not controlling and generally of limited suasive value.
If someone cites a bogus case, wouldn’t that be obvious to the opposing counsel/judge? Isn’t software readily available to check citations? As an attorney, I’m not sure what I would want more than a ready opportunity to say, “My opponent is a lying sack of shit!” (Makes me wonder why firms would not have their own pleadings run through such software.)
IMO, the practice of law is becoming such a numbers business. Doing as little as possible, throwing shit against the wall, hoping to make a profit out of the numbers. Or the alternative, trying to bury a client’s culpability - and one’s opponent, in unnecessary paperwork and verbiage. The disreputable use of AI is just one of many facets of law as it is currently practiced.
Of course, one can argue that the asshole atty’s client ought not be punished for their lawyer’s transgressions. I think that somewhat murky. IMO, a client has the opportunity to hire a non-asshole as easily as an asshole. I think there should be some public notice to allow a potential client to know that this lawyer’s transgressions has caused their clients to lose cases in the past. I’d probably support a “3-strikes” policy, with increasing fines and suspensions. Of course, it isn’t as tho we have too few attorneys, so I’m not heartbroken about the idea of disbarring several.
Final point, IME, Rule 11 sanctions are not invoked with any consistency, such that their imposition often impresses me as almost arbitrary.
Or, perhaps more likely, one of their clerks.
The court actually did include a mild admonishment of opposing counsel for failing to catch the discrepancies themselves.
It’s not just lawyers.
The Newfoundland & Labrador government commissioned a report from Deloitte on projected improvements to the health care system.
Cost $1.6 million.
It has fake citations.
Deloitte says it doesn’t matter. They stand by their recommendations.
https://www.cbc.ca/news/canada/newfoundland-labrador/wakeham-health-report-sources-9.6992094
Right, because naturally a 25 year old clerk with no legal experience is a better writer than a mature judge who has spent their career writing briefs, reports and judgments.
IME, it really varies from judge to judge. Some few write most of their own stuff. Others, it used to be almost a joke when the judge’s decisions would really change and you’d figure, “Guess someone got a new clerk.”
IME, the most common is for clerks (or staff writers, who can be well older than 25) to draft an opinion, which the judge edits.
Why do you think judges have clerks? And why do you think being on the law review is almost (practically) a requirement to be a clerk in the US, especially at the federal and appellate level?
Whether a judge is themselves a great writer or not, you can pretty much guarantee the first draft was written by a clerk, to the judge’s direction as to overall theme/conclusion.
I’m well aware of why judges have clerks, thanks.
Hey, for all I know Canada is different. You seemed incredulous at the possibility a law clerk wrote the opinion.