Will AI take big law litigators’ jobs? Forum

[crossexamination]

Assume within 5 years, AI will be able to (1) do doc review more accurately than a human associate; and (2) do more comprehensive and precise legal research and writing than a human associate.

Will there be mass layoffs (60-80% reduction) among big law litigation associates? If not, what will be left for them to do?

Will billing rates go down dramatically among lit associates and partners once AI is smarter and wiser than them?

[JorgeMichael]

If you take the (wildly unrealistic) assumptions you've made as true, then maybe. Ultimately firms will continue to hire junior associates as long as clients are willing to pay for their time. Software that can do legal research and review documents likely can't be billed out in the same method as juniors can, and likely it would be more profitable to just have the juniors.

[Anonymous User]

You're still going to need a talented lawyer to direct the AI and figure out the strategy. Even with your assumptions, it's not like you can hand a computer a copy of the local rules, the documents that could be produced in the case, the documents you receive from the other side, and a Westlaw login and let the computer figure everything else out. You'll still need attorneys to interview/interact with folks at the client, pick witnesses, develop case themes/strategy, decide what discovery you want from the other side and what discovery you're willing to give, prepare for/take/defend depos, prepare for trial, consider what a jury would want, etc. The list goes on, so there will still be plenty of things for associates to do even if we can pass off the initial doc review, research, and writing to a computer.

What you're more likely to see is the continuation of a trend that started in 2008 - clients don't want to pay for menial tasks like doc review, so firms have positioned themselves to capitalize on the higher level strategic thinking that computers (or cheaper doc review outfits) can't do.

I'm an associate in a well known IP lit practice (hence anon), and doc review and legal research/writing are a very small portion of what I do every day. I'd be glad to hand those tasks off to a computer and continue doing the more interesting stuff.

[Anonymous User]

If AI becomes advanced enough that it can do more precise, comprehensive, and coherent legal research and writing than a human, what jobs would be safe? Am I to assume that such advanced AI couldn’t also do transactional work? But I do think some of the more menial tasks are close to being handed off to AI.

[nealric]

I think it will be a very long time (if ever) before a client would be willing to take AI-generated product in a raw and unreviewed state. If you got to that point that was something people were generally willing to do, why bother with the formalities of litigation at all? Just feed the relevant documents into the AI judge and it spits out a final result.

Setting that aside, there's no reason why litigation would be impacted more than other roles. A lot of junior corporate work is just reviewing diligence (not much different than doc review as a cognitive task) or taking form documents and conforming them to the relevant transaction. There's actually quite a bit of potential for existing AI to do a lot of junior corporate work. It's only a matter of time before specialized legal AI is commonplace.

Still, I think the result is going to be more akin to what happened when accountants went from a paper to computer spreadsheets. It used to be junior accountants spent a lot of time doing rote arithmetic. Basic doc review or replacing names and dates on a form will be looked at like doing manual arithmetic at an accounting firm.

[Anonymous User]

Is the prevailing view really that there will be no reduction in lit associates as a result of AI?

Doesn’t it seem more likely that junior or midlevel work will be significantly reduced and the need for senior associate + partner insights will remain constant?

[Anonymous User]

Why wouldn’t firms take on more work to compensate, and just give junior/midlevels slightly different work?

But in lit, I can’t really see AI generating useful pleadings, or even that many helpful responses to discovery. At least, not in any kind of practice that isn’t already using stock everything. As the third post here says, you’re still going to need someone to identify/compile the materials for AI to use, and figure out what you need it do. (Leaving aside potential security concerns.)

The idea that AI will eliminate the need for lawyers makes me think of how a family member described one of their PhD profs back in the 1990s, when the internet was new and when you started to get digitization and online access to stuff and so on. This prof was convinced that this was the wave of the future and kept coming up with bizarro/infeasible plans for his doctoral students, because “computers can do it!” Spoiler: computers could not, in fact, do it.

[nealric]

Is the prevailing view really that there will be no reduction in lit associates as a result of AI?

Doesn’t it seem more likely that junior or midlevel work will be significantly reduced and the need for senior associate + partner insights will remain constant?

I think biglaw as a whole already takes fewer junior associates than it did 20 years ago. The structure is less pyramidal and less "up or out" than it was then. But there's a limit to how many juniors you can eliminate. Senior associates + partners aren't immaculately conceived. They have to put time is as juniors before they can become senior. Tech has (and will continue) to eliminate lower-end tasks.

Tech has also already decimated the ranks of support staff. When I started in biglaw, there was one secretary for every four lawyers (typically 1-2 partners plus 2-3 associates). There were even a few who had personal secretaries that dictated and read all of their correspondence. By the time I left biglaw it was more like 1:12 with just a secretary pool. AI could nearly eliminate legal secretaries altogether. It will also lessen the need for paralegals by automating things like cite checking.

[Tide030]

AI appears likely to change big firm staffing, recruiting, and compensation. A study just came out noting Lexis’s AI tool had about 20% errors. If the rate of current improvement continues, I imagine a 95%+ accuracy is likely in 5-10 years for doc review and motion practice. Firms will still need senior associates and junior partners to review that work, though. Which makes junior associate hiring interesting. Firms won’t need as many of them, but will need some to move up the ranks (and review AI-generated work). Also seems like clients won’t want to pay to train junior associates on doc review and motion practice if AI could do it instantly—encouraging non-billable-hour payment methods. I could see big firms hiring more clerks and high-level and creative thinkers with strong soft skills but less grinders or research/writing-types who don’t want face time with opposing counsel, clients, judges/juries.

[Anonymous User]

AI appears likely to change big firm staffing, recruiting, and compensation. A study just came out noting Lexis’s AI tool had about 20% errors. If the rate of current improvement continues, I imagine a 95%+ accuracy is likely in 5-10 years for doc review and motion practice. Firms will still need senior associates and junior partners to review that work, though. Which makes junior associate hiring interesting. Firms won’t need as many of them, but will need some to move up the ranks (and review AI-generated work). Also seems like clients won’t want to pay to train junior associates on doc review and motion practice if AI could do it instantly—encouraging non-billable-hour payment methods. I could see big firms hiring more clerks and high-level and creative thinkers with strong soft skills but less grinders or research/writing-types who don’t want face time with opposing counsel, clients, judges/juries.

I wonder if this means that either (a) hiring becomes even more prestige-oriented as the grinders from lower-tier schools get culled or (b) hiring becomes way more personality-oriented so that all grinders lose out equally.

In a reasonable world it should probably be (b), but (a) is probably more in line with how current partners are thinking about the issue.

[Anonymous User]

AI appears likely to change big firm staffing, recruiting, and compensation. A study just came out noting Lexis’s AI tool had about 20% errors. If the rate of current improvement continues, I imagine a 95%+ accuracy is likely in 5-10 years for doc review and motion practice. Firms will still need senior associates and junior partners to review that work, though. Which makes junior associate hiring interesting. Firms won’t need as many of them, but will need some to move up the ranks (and review AI-generated work). Also seems like clients won’t want to pay to train junior associates on doc review and motion practice if AI could do it instantly—encouraging non-billable-hour payment methods. I could see big firms hiring more clerks and high-level and creative thinkers with strong soft skills but less grinders or research/writing-types who don’t want face time with opposing counsel, clients, judges/juries.

I wonder if this means that either (a) hiring becomes even more prestige-oriented as the grinders from lower-tier schools get culled or (b) hiring becomes way more personality-oriented so that all grinders lose out equally.

In a reasonable world it should probably be (b), but (a) is probably more in line with how current partners are thinking about the issue.

Yes, strong favor towards prestige, at least right out of school. But lawyers lateral often now. So I could see a decreased emphasis on summer associate recruitment and a stronger emphasis on hiring clerks and laterals with proven experience. New associates will no longer be the valuable revenue generators they are now. Why hire as many first years and invest in their development if you can’t really profit on them and they will likely quit/lateral in a few years? AI will replace the bottom layer of pyramid. And teaming up with specialist boutiques, I think, will become more common. Like when Beth Wilkinson handled the preliminary injunction hearing for the FTC’s challenge of the Microsoft-Activision acquisition, while Sidley and Skadden did all the heavy lifting behind the scenes.

[nealric]

AI appears likely to change big firm staffing, recruiting, and compensation. A study just came out noting Lexis’s AI tool had about 20% errors. If the rate of current improvement continues, I imagine a 95%+ accuracy is likely in 5-10 years for doc review and motion practice. Firms will still need senior associates and junior partners to review that work, though. Which makes junior associate hiring interesting. Firms won’t need as many of them, but will need some to move up the ranks (and review AI-generated work). Also seems like clients won’t want to pay to train junior associates on doc review and motion practice if AI could do it instantly—encouraging non-billable-hour payment methods. I could see big firms hiring more clerks and high-level and creative thinkers with strong soft skills but less grinders or research/writing-types who don’t want face time with opposing counsel, clients, judges/juries.

I wonder if this means that either (a) hiring becomes even more prestige-oriented as the grinders from lower-tier schools get culled or (b) hiring becomes way more personality-oriented so that all grinders lose out equally.

In a reasonable world it should probably be (b), but (a) is probably more in line with how current partners are thinking about the issue.

Yes, strong favor towards prestige, at least right out of school. But lawyers lateral often now. So I could see a decreased emphasis on summer associate recruitment and a stronger emphasis on hiring clerks and laterals with proven experience. New associates will no longer be the valuable revenue generators they are now. Why hire as many first years and invest in their development if you can’t really profit on them and they will likely quit/lateral in a few years? AI will replace the bottom layer of pyramid. And teaming up with specialist boutiques, I think, will become more common. Like when Beth Wilkinson handled the preliminary injunction hearing for the FTC’s challenge of the Microsoft-Activision acquisition, while Sidley and Skadden did all the heavy lifting behind the scenes.

They will continue to hire juniors for the same reason they do now: it's cheaper than hiring a seasoned person. As long as you can bill them out, it doesn't matter if they add all that much value at the beginning. As a client, I've had plenty of experiences of having to have a junior on a call to "take notes" or other makework task. It's really just a way of charging more for the partner's time. You can't access the $2,000/hr Sr. partner without also speaking to their $1,500hr Jr. Partner and $700/hr Jr. Associate, so it really costs $4,200/hr to talk to the Sr. Partner. It's just in the business of law, every attorney is a line item on the bill. In other businesses, training gets baked into other costs.

Highly desirable partners only lateral if they are paid big bucks to do so. Plus, experienced lawyers have to come from somewhere. However, I think we are already seeing a bit less "up or out" model and more focus on retention. The traditional model of hiring 50 summers to make 5 partners in 10 years is probably out in favor of trying to hire 25 summers to make 5 partners.

[Anonymous User]

AI appears likely to change big firm staffing, recruiting, and compensation. A study just came out noting Lexis’s AI tool had about 20% errors. If the rate of current improvement continues, I imagine a 95%+ accuracy is likely in 5-10 years for doc review and motion practice. Firms will still need senior associates and junior partners to review that work, though. Which makes junior associate hiring interesting. Firms won’t need as many of them, but will need some to move up the ranks (and review AI-generated work). Also seems like clients won’t want to pay to train junior associates on doc review and motion practice if AI could do it instantly—encouraging non-billable-hour payment methods. I could see big firms hiring more clerks and high-level and creative thinkers with strong soft skills but less grinders or research/writing-types who don’t want face time with opposing counsel, clients, judges/juries.

I wonder if this means that either (a) hiring becomes even more prestige-oriented as the grinders from lower-tier schools get culled or (b) hiring becomes way more personality-oriented so that all grinders lose out equally.

In a reasonable world it should probably be (b), but (a) is probably more in line with how current partners are thinking about the issue.

Yes, strong favor towards prestige, at least right out of school. But lawyers lateral often now. So I could see a decreased emphasis on summer associate recruitment and a stronger emphasis on hiring clerks and laterals with proven experience. New associates will no longer be the valuable revenue generators they are now. Why hire as many first years and invest in their development if you can’t really profit on them and they will likely quit/lateral in a few years? AI will replace the bottom layer of pyramid. And teaming up with specialist boutiques, I think, will become more common. Like when Beth Wilkinson handled the preliminary injunction hearing for the FTC’s challenge of the Microsoft-Activision acquisition, while Sidley and Skadden did all the heavy lifting behind the scenes.

They will continue to hire juniors for the same reason they do now: it's cheaper than hiring a seasoned person. As long as you can bill them out, it doesn't matter if they add all that much value at the beginning. As a client, I've had plenty of experiences of having to have a junior on a call to "take notes" or other makework task. It's really just a way of charging more for the partner's time. You can't access the $2,000/hr Sr. partner without also speaking to their $1,500hr Jr. Partner and $700/hr Jr. Associate, so it really costs $4,200/hr to talk to the Sr. Partner. It's just in the business of law, every attorney is a line item on the bill. In other businesses, training gets baked into other costs.

Highly desirable partners only lateral if they are paid big bucks to do so. Plus, experienced lawyers have to come from somewhere. However, I think we are already seeing a bit less "up or out" model and more focus on retention. The traditional model of hiring 50 summers to make 5 partners in 10 years is probably out in favor of trying to hire 25 summers to make 5 partners.

Sure. That makes sense to me. So how will the market respond to a 50% reduction on first-year associate hiring? I’d assume the prestige of these positions (and getting to work with an elite litigation or trial partner) will increase. But the overall need for fewer biglaw associates will suppress associate wages generally, especially in lower-tier biglaw and midlaw firms, which will have a huge labor pool to hire from. I also wonder whether partners will play a greater role in selecting/hiring the associates they want to work with.