Ethical AI and Legal Innovation with Sarah Alt

Thank you, Daniel. Thanks, everyone. It's great to be here.

I've been a technologist my entire career, and I was definitely one of those that, you know, I've been alongside technology advancements, you know, my entire career. So, you know, to me, especially when generative AI was exposed to the world and we had access to it in November of twenty twenty two, it was like, okay, finally, like, people are starting to really see how you can use this every day. Meanwhile, many of us knew about, you know, artificial intelligence and machine learning and advanced automation, and those things were not new. It's just the way in which we were able to release this for generative AI was just so remarkable.

So, you know, for me, my expertise is really has been more on the ethical and responsible use. It's been my entire career, my mission has been to explain, how technology works. It's just always what I've been good at being able to do. And so for me, this was just another chance, another opportunity to do that with such an awesome technology.

Yeah. So my moment, my kinda catalyst trigger moment was, when I was in house in an IT function, I was responsible for bringing in all the new technologies and the investment portfolio of new technologies.

We were hiring a manager to join our team, and we posted the job for like two weeks. And you know, the batch of resumes, and we got the candidates that we got. We looked through that batch of resumes, and it just we didn't quite see what we were looking for in that batch of resumes. And so we said to our HR business partner, I said, hey.

Go back and see if you can get another round of applicants. Right? And I fully expected that she would have to post the position for another two weeks. We'd get all of these another round of candidates.

She came back that afternoon, like, two hours later with another batch of candidates. In that batch of candidates, we did find somebody that, you know, kinda matched what we were looking for, and we went through and proceeded with the process. But then I in a follow-up, I asked her, I said, so where did this other batch of candidates come from? Because you clearly didn't repost the position.

Right? She said, well, these were the ones that were originally screened out.

And I said, screened out by whom? You know, like, so I'm thinking there was a human deciding that, well, this is the batch that Sarah is gonna we're gonna take to Sarah first, and then we'll come back with another bet. And she said, well, they were screened out for the system. And I said, well, how did the system know, like, this time that it was wrong?

Like, it wasn't what I was looking for and I would have missed these candidates. And the more I asked that question or dove and tried to sort of untangle that because remember I mentioned earlier that, like, my mission is to explain how technology works, and here I'm sitting as a, you know, technology leader who couldn't have the technology explained even to me. So then how could I be a good representative of this technology to anybody else in the company at the time? And the more I asked her, the more we realized we needed to get the vendor at the table to try to explain that technology.

Once again though, we were asking questions that either and I could sense that we were either headed down a path where the provider didn't have the answer or wouldn't give us the answer because at some point that might reveal trade secrets or other intellectual property or whatever it is about how that technology works, but it's possible they also didn't entirely know how that automation works. Now what we were using at that time is a very common, you know, system in HR, common HRIS type of application or human capital management application called applicant tracking system, ATS.

This was not back when artificial intelligence, was really being talked about in the workplace. So I would not have categorized this as an AI application the way that we understand and know and are qualifying that statement today. But that doesn't matter because the point is it was an automation system.

And so that event, that kinda moment, that sort of just discomfort that I had with how is this working and could I explain this, led me down sort of this deeper path and rabbit hole of research to realize that, you know, where we're headed with automation, where we're headed with algorithms, where we're headed with how these systems work, especially in that context, which is, you know, hiring or other, you know, talent management types of systems, It just made me uncomfortable. And so I, you know, set myself on a mission to just say, hey, while I'm working, while while I'm working as a technology leader, I'm going to stay on this mission to make sure our profession does better by this technology, can explain it, and does understand where the impact might be. And that eventually led me to starting my own consulting practice.

I left that company, started my own consulting practice, and then as I was getting clients who were AI startups at around that time, so this is, like, two thousand eighteen, two thousand nineteen, and then I realized that, like, there not only was a need for this kind of consulting for clients, but also just, again, to serve that mission for me, I basically split, sort of my business model from the for profit work that I was doing and then the nonprofit work, which is when I started up the Ethical AI Consortium, was established so that I could basically be, you know, a chief explanation and engineer for everyone, for society, and give that awareness and explanation away as part of my mission, fully knowing that I had the for profit and the ability to generate revenue and the consulting for my clients that actually needed to have responsible products and services.

I love that. So you're so visceral. It's such a great word because, yes, it very much did feel like that. I certainly felt like I was challenging a vendor for the first time in that domain of topics, right, in the way I was asking those questions. And so, yes, I mean, I did feel like a lone voice in the room.

And but at the same time, I also could tell from the way that I was approaching this thinking about, you know, labor and employment law before I really, you know, worked for a law firm clearly. But my internal client was the HR business partner in the room. And so I feel that she also felt somebody was empathetic to her situation, that she had never asked these questions before, and she was realizing that, oh my gosh. Maybe I need to start asking these questions differently because Sarah is not the only person who buys technology in the company.

I'm also, as an HR person or any other, you know, business functions, maybe people who are, if not the buyer themselves, they're certainly the decision makers or the users of it. You know? So I did feel like I was a lone voice. But again to me it just felt right.

It was the right thing to do. I had the blessing of having worked for cultures in my past that were very much focused on cultural values and on commitment to, you know, always putting employees first. Obviously, following the law is what every company has to do when it comes to bias and discrimination.

And to be clear, speaking of that, I was not at all suggesting that we may have been discriminating or biased in how we worked. But my point in asking these questions was, I didn't know that we weren't. We certainly weren't doing it intentionally, but we could have been without knowing it. And that's really kind of where I really started getting, more passionate about and eventually led to the EAIC was I wasn't as necessarily as concerned about the intended consequences. I mean, of course, intended consequences of anything are bad. I was worried more about trying to explain the unintended consequences.

These things like, if you don't ask the right questions or if you just assume that technology or the system is smarter than humans or is better than humans or more sophisticated than humans, Like, you could get into some really tricky situations with unintended consequences because you couldn't explain yourselves, you couldn't explain the technology. That's really what drove me because I was confident that, you know, labor and employment laws would guide whether we were doing anything intentionally.

I'm more concerned, and as I've always been with my clients, about making sure they understand the unintended consequences.

Yes, sir.

I know. I know. Right? It's like such a serendipitous combination of all these things. Right? So what first got me into working with law firms was through my strategic consulting. It was because I had such a great relationship with my outside counsel when I was working for those for profit entities.

And we had a really great way of organizing chaos and turning it into a machine and a very great strong operation on just how to get all of the legal needs that we had and all of the matters and all the contracts and everything that we were doing. We just had a really great process and a really great system. So she called me up and said, hey, I know you're you're doing your own consulting thing. How about coming to my practice?

Help me figure out how we can reproduce what you and I had at that for profit entity. Let's do that for some of my clients that are very similar. I said, sure. Super.

So that's what got me just, like, really fascinated with, you know, frankly, the operation of and the machine of the practice of law.

The other the serendipitous thing is when you're talking about ethical and responsible use of anything, much less artificial intelligence, I felt like I was amongst my people. People who live every day based on an ethical code. Their license is taken away if they violate their ethical duties. Right? Not many professions, certainly none that I would have been in the sectors that I had worked in. So that's what kinda just got sort of the ball moving.

When I started working with Michael Best as one of my strategic consulting clients, I approached the at the time, Jason Schultz, our chief innovation technology officer. He had a position open as Director of Technology Integration and it was right up my alley with some of the work that I was doing with them and I said, but this is really interesting to me, but I would still want to be able to do my ethical AI consortium work. That's my mission. That's who I am.

That was not at all a problem for Michael Best because we're not just a law firm, we're also a consulting and strategies practice. We're very nimble from that standpoint. This whole idea of somebody coming in, a staff member coming in and being able to also be billable to clients and have an internal responsibility firm, but also have whether it's a mission driven or a revenue driven opportunity to consult with clients was a no brainer for Michael Best. Other firms, you know, they said this wouldn't work here.

Right? And I get that. I understand because they may not have the flexibility to do that through the strategies and consulting business. Again, it was like all these things kind of serendipitously coming together, and that and that's what landed me at Michael Best.

Yeah. So we definitely have been already entertaining discussions about client confidentiality and privilege. I think that certainly is the number one place in terms of, like, the ethical impact.

More conservative clients will specify, you know, that we cannot put their data into AI without their permission. We're seeing that already in our outside counsel guidelines. Others other clients may be asking us to you actually use AI when we have vetted the privacy and confidentiality and confirm that it would be beneficial for us to use it for serving them. So we're we're seeing just like this spectrum of clients, but, again, we're used to that because, you know, as a corporate firm, we have all kinds of different clients on various different risks and trust perspectives.

So I do believe personally that this is kind of the time that we're gonna start seeing this maybe settle into a better rhythm of trust and confidence as more and more clients continue to, you know, start getting comfortable with artificial intelligence and using artificial intelligence. I also believe that as the barrier cost to entry for more legal technology, like, lowers as that barrier you know, that cost to entry to, how you can even get access to the technology as that lowers, I think we'll start to see more standards, more commonly accepted technologies or ways of working. It's just that it's so, like, wild right now in terms of everybody trying to get in, and some of this is very niche.

Right? Some of these technologies are very niche, And so they may do one thing really well, but those smaller, shorter putts may not be worth it until, you know, the technology matures a little bit better.

Yeah. For sure. I think the valuation, is really important. And we know this from all kinds of artificial intelligence, you know, innovation and products and services. So, yes, I think that is certainly a priority for us is making sure that we are able to identify and tag and start giving even more of the data that we have, like, the identities that we need in order to make it useful for us in artificial intelligence. I'll use just two general examples. Like, if we are producing let's say we're producing handbooks for clients, like we have a lot of handbooks, employee handbooks that we produce for clients.

And, you know, being able to locate all of those so that we could, for example, scan those and confirm, you know, which ones need to be updated for our clients, which ones do we have a new way of drafting an attendance policy or or whatever that might be that really makes Michael Best awesome. That is certainly part of or an essential ingredient of the legal work that we do. It's a knowledge profession, and so it's based on that knowledge and expertise of all of the talent that you have in your firm, and that talent is producing all that great data.

Being able to harness that, being able to look at that. The second example I'll give is even just internally for when we have new associates or when we have paralegals that are learning a new specific trade or a new specific type of document or way of documenting or even, you know, some of our classical writers, like the writers that you turn to in your firm where you're like, that's the person that should write this because you know the situation that you or your client are in. So we wanna make sure that we can harness that style and that way of writing and make sure that we can impart that on other people. Right? So, you know, before we had artificial intelligence or before we had the power of technology, that would usually be done side by side. That would be done, you know, in your mentoring opportunities.

And now we, you know, have opportunities to, use artificial intelligence to get us there. And to do that in a way that once again, we can say, don't just draft this the way a law firm would do it. Draft it the way Michael Best would.

Yeah. So, I'm not at liberty to give my opinions about regulation or legislation, but I can respond with what what we know has happened already. So since GDPR and CCPA, we've observed a time when data privacy certainly came first out of necessity. And it sort of makes sense because, like we talked about, it's an essential data is an essential ingredient before anything can be baked into AI anyways. So AI legislation has been slowly following.

But more recently, we've observed states like Connecticut where their data privacy bill packaged AI together, sort of the why wait for two bills, let's start moving this forward as one, even at the federal level. So the most recent executive order authorizes the attorney general to prevent large scale transfer of our personal data to countries of concern and provides guidance for and safeguards around, you know, the activities that would give countries access to our sensitive data. So if you couple that most recent executive order with the executive order from October of twenty twenty three, the executive order on AI, we can start to see how these governing models and these governing approaches and the various agencies that are given those orders from that executive order are going to be intertwined, and we're gonna see data privacy and artificial intelligence, I think, travel more together than we had seen before.

Yeah.

In order to educate the lawyers, they need to be able to understand, I would imagine, Sarah, one, what impact does it have on me? Mhmm. Two, what impact does it have on my clients? So there's a lot going on, as Jared noted, right, with respect to regulations. You just, elucidated on various different regulations that are already in place.

Yeah. For sure. So very early on when we appointed when the firm appointed me as the, chief AI officer, we established an AI center of excellence in the firm. And it started out as a group of fifteen practitioners from all of our major disciplines of law, areas of law, in recognition of the fact that the way Michael Best approached this was AI is going to disrupt both the practice of law in every area of law, but it will also disrupt meaning with, as we've already talked about, the technologies you're going to be used, how you get work done, so there was that internal how it's going to disrupt you, but then also it's going to disrupt all of the areas of law in terms of bringing new opportunities because artificial intelligence has all of this powerful capability to influence the way, you know, our clients go about conducting business or serving their customers and their clients.

Right? So we started with our AI center of excellence to just, you know, really have that open discussion and conversation and that creating that awareness, demystifying, we called our first webinar for them the demystifying artificial intelligence and somewhat, again, creating that sense of awareness and urgency to say that this is not just for the technology attorneys. This is not just for the AI attorneys, which, by the way, there weren't any. There aren't you know, like, there this is so new that it's a matter of where do we need to guide you and direct you to getting yourselves the right research and the right access to the information so that you can, you know, prepare for and be ready for this kind of change.

I'm most excited for how AI can help lawyers cut down on the administrative burden and keeps them being awesome at what they do.

Burnout in this profession is still high. When I'm with high school students talking about a career at a law firm, the ones with parents or family members who are attorneys will say to me, but what about the burnout and the stress? And this is from high school students. So they see it.

They know that this profession still kind of has, a little bit of that reputation to shed and or many, many people just accepted it. Right? And so It's one of my missions as I continue to serve our practitioners in the firm is to frankly do whatever we can to rid that burnout by enabling super technologies and just try to restore this excitement in the practice of law. If not excitement in the practice of law, at least that improving the quality of life, improving the work life balance in the way that we work.

Because again, in a human practice, like in a knowledge profession and the importance of humans being available, and if you're not there for your clients, you can't serve them. And with that such high regard for your duty to serve, if you're not able to put the oxygen mask on yourself and ready yourself and be in the right place, you're not there for your clients.

Burnout in this profession means then that your clients don't have you at your best self. And I just hope that, you know, anything I can do in bringing artificial intelligence and technologies to help improve that quality of work will improve their lives as well.

Thank you, Daniel. Thank you, Jared.