Related Posts
Most sweatshop agencies?
Visual Storyteller
And now.. let me take a selfie!
Conversation Starter
Visual Storyteller
Additional Posts in Law
How do you wind down after a rough day?
New to Fishbowl?
Download the Fishbowl app to
unlock all discussions on Fishbowl.
unlock all discussions on Fishbowl.
I am wary of it as well but have started to experiment with it. It still needs a lot of human support in my experience but does make some things easier.
What things does it help the most with in your experience?
I get the ethical hesitation, but honestly? We’re hitting the point where not using AI creates its own ethical wrinkle. It’s not just *“should lawyers use new tech,”* it’s *“should clients have to pay triple because we refused to?”* If I can have AI generate a first-pass structure, issue map, and case-law vector in 30 seconds, and then I spend my actual time doing the judgment work, that’s a net win for everyone. Imagine telling a client, “Yes, we spent four hours drafting this memo because I didn’t want to let a machine outline it in 10 seconds.” The tech isn’t replacing legal reasoning — it’s just eliminating the parts of the workflow that used to be a time sink. So the question isn’t just whether AI is ethically acceptable; it’s whether refusing it is.
Perfect answer, completely on point 💁♀️
It's getting to the point that it will be unethical not to use it because you'll be wasting client time/money. Like using volumes of books instead of using Google.
Not saying it's perfect, but .2 + another .2 of correcting it is better than 1.4 of old fashioned research.
If this is true then why don’t law firms lower their rates?
You literally have to use it or you’ll be the boomers who wouldn’t use email
I am a huge AI hater and my firm tells us not to use it. It think it may have a use in research and like doc review or deposition summaries. If I were a judge and someone submitted an AI motion, I would be furious.
It seems to be helping pro bono types create multiple pages of garbage with cites to irrelevant case law. It can’t discern between holding and dicta
You are not alone! I am very against AI for many reasons - environmental impacts, impacts on communities, the creation/consolidation and dominance of the market by a very few companies. The truth is it is being forced down our throats by venture capital whether it helps us or not, and this is the time we should actively resist allowing it to become a norm. It's not a norm yet and doesn't have to be.
In response to the comment about whether a small group actively resisting will make a difference, no, I don't think such a group (including myself, at least at the time being) is likely to make a meaningful difference. If anything slows down the ongoing adoption of AI, it will likely be through something that happens, or fails to happen, with respect to AI itself, such as AI fails to decrease the number of AI hallucinations and continues to be unreliable in professional settings due to that and other risks.
However, for me, if I have ethical qualms about something, I try not to make a decision about whether to use the tool based on whether I will make a difference, but whether I will feel like I made the right choice, which might even involve using that tool.
I order about 80% less than I did three years ago off Amazon. It's not going to make any difference to Amazon's bottom line and may cause me more minor inconvenience with higher costs or effort to buy most items elsewhere, but I'm happy with my choice.
I don't mean to suggest that this is the superior way to think about these issues, but for me, I can't say yes to your question. However, it won't impact my ongoing decisions about whether to use AI.
I’ve been using Harvey (which itself uses gpt, Claude, etc) some recently to do quick-and-dirty analysis of a few hundred employment contracts as part of deal diligence, which has been fascinating. The contracts are closely based on the same form and largely use the same language. About 20% of the time, it interprets the same exact assignment language (so nothing even particularly complicated) differently. One day it decided it felt like charting some contract analysis in Spanish (all contracts are exclusively in English). This technology simply is not built to be accurate, so ask yourself when using AI: is this a task that only needs to be 80% accurate?
My firm has at least one very large client who requires outside counsel to use AI for factual summaries like deposition transcripts, court decisions, and complaints. They’re not going to pay an associate to sit and write that when AI can do it and the associate just has to review and tweak it.
Interesting re: summaries. The legal research—even on Westlaw AI—is spotty, at best. It always positively answers my question and then the case won’t typically stand for what it suggests. I’ll then reverse the question, seeking the opposite conclusion, and it will find cases supporting the new question.
I’ve seen AI motions filed by attorneys that were so clearly AI-drafted. So. Many. Pages. So. Many. Cases. They’re exhausting to counter—like getting papered to death. I can only imagine how these attorneys are billing their clients.
I’m not against the use of AI but I do feel there’s a lot of gray area that is going to come up in the future. I’m in healthcare and everybody has some new AI health thing they want and I have a fear there’s going to be a lot of privacy risks that are going to come up. We try to add safeguards to the contract but it just seems like there’s so much we don’t fully understand and may be missing
This. I feel like it has exploded way too quickly. We don’t know the actual implications
It’s gonna halve the number of lawyers required over the next 5-10 years and you’re worried about the ethics of it?
Yes I am, AGC1. I’m very concerned about the environmental impacts and the ethics of using something we know is damaging to the planet and negatively impacting lower income communities. It’s concerning to me that that isn’t a part of the conversation
Use it to spot issues; draft initial correspondence on matters, etc. You still have to feed it information for it to be useful.
The American bar Association has already issued a formal opinion on it I believe and I’m really looking forward to their next one. I don’t understand why this is still a question when we already have guidance. Yes we need to be careful. Yes we need to be ethical. But it is a tool at our disposal, and I think it’s our responsibility to use any tool at our disposal to help a client.
I’d argue competency in tech doesn’t require consistent use or reliance on it
I'm not a fan of it and use it rarely. I was reading about Luddites and I suppose I fall into that camp. Great discussion!
Made a similar post on this topic a month or so ago. I use AI as a way to jumpstart research or significantly streamline contract review, and think it's great as a way to save time on those types of tasks (frankly, the stuff a first-year or junior associate would do). Some attorneys definitely over-rely on AI tools without bothering to learn the underlying issues, and it shows.
Do you use ChatGPT or do you recommend some other platform?
Mediation statements - holy cow AI is great for those. Also drafting hard hitting intros to a brief you just finished the old school way - another killer AI use case. I should stop sharing my trade secrets
Corporate, lev fin, real estate, IP, litigation discovery, bankruptcy
I’ve been holding out too, but I fear it’s not going anywhere
Great 🫠
AI is inevitable. You will be left behind if you don’t appreciate it and wield its power for good. Imagine saying in the dot com era or 2000’s, “the internet is unethical and I refuse to use it.”
I don’t think that’s a fair comparison. AI has far greater environmental implications than the internet did, and we’re fully aware of them. Also, we are capable of doing all the things AI does, it seems it’s just being pushed to get more done with less brain power. In a field where we’re literally counting the minutes we work it’s just going to contribute to more unrealistic expectations and standards.
Not to mention it wrong a LOT and fabricates stuff and then you have to take time figuring out whether what it told you was real. No thanks. Stop forcing that 💩 on me in my PDF and email previews.
Not alone. Publishing a book on this soon. Been writing it since before generative AI, Just haven't been able to keep up with changes since LLMs became a thing. Gonna publish anyway.
I’d love to read it when it’s finished