John Albright. Courtesy photo.

As the chief legal officer of Chicago-based HUB International, an insurance brokerage company, John Albright deals with tons of data. This reality has inspired him to adopt new technologies, including what he calls “AI Lite.”

The tech-minded GC sees many more opportunities to use AI down the road across the whole legal industry.

Albright recently chatted with Corporate Counsel about what his company is doing with AI now and how new tools could affect legal departments in the future. This interview has been edited for clarity and length.

Corporate Counsel: What are some of the legal and regulatory challenges you deal with as the general counsel of an insurance company?

John Albright: We're on the brokerage side so we're less regulated than carriers, but we deal with a lot of the same issues that any large company would have. First and foremost is data privacy and security. We're a large employee benefits broker so HIPAA is front and center. Beyond that, insurance is regulated at the state level. So New York, for example, just came out with state-specific cyber regulations that all insurance and financial services companies have to comply with.

When did you first notice artificial intelligence being used in legal departments?

I think it depends on how you define AI. It depends on whether you include document creation tools and document management platforms, which have been around for years, and have some aspects of machine learning and automation in them. The challenge with most of those platforms to this day is that they're not constantly learning and making adjustments.

If you're looking at true AI where it is constantly making adjustments and learning from what happened today, that's far more recent. I think most people in legal technology take the broader definition of it and those tools have been around for years.

Does your office use what you would consider “true AI”?

We use “AI Lite,” which is the more static version of it. We've used a policy comparison tool for years which does automatic data extraction and then takes that data and puts it into a quickly readable template regardless of where it falls in a document. We also have a form generation tool that we've been using for a number of years. It's debatable whether or not that is true AI, but that's where we are. We are looking at expanding into a more advanced version of AI.

How long have you been exploring the use of a more advanced AI?

Probably within the last year in earnest. We've been looking at it on the litigation side even more recently. The broader application is an advanced AI overlay on our document management platform which would facilitate data extraction, analytics and things of that nature.

Is there any hesitation on your part in incorporating more advanced AI in your office?

Anytime you're looking at large amounts of data there is a significant up-front investment and you're kind of married to a platform on some level. We are going to do some long testing to make sure we have the platform that can do what we need it to do today, but also is in a position to evolve as technology advancements and data collection expand. We see these as living tools.

What are some of your concerns when it comes to AI?

You have to understand the limitations on it. If you look at the advancements over the past couple of years they have been impressive by any measure. AI is still not making meaningful judgment calls. Any of its output is still going to require a fair amount of human oversight.

Do you think AI technology could eliminate jobs from attorneys or paralegals?

It is going to eliminate certain jobs. It may eliminate the need for three paralegals, you need only one with it. I don't know what that looks like. If you look at what AI is doing currently, not specific to legal, but the average employee making $200,000 a year—which is a senior-level person—spends about 30 percent of their time doing things that AI can do today. That's a pretty significant statistic. It's impractical to say that you could bring in AI tools to eliminate all 30 percent of that. But there is a lot that AI can do that more junior people in the legal world are currently doing.

There will be fewer people delivering the services that you get—but the people are not going away.

What should legal operations directors or heads of legal departments have on their mind when going to purchase AI technology?

AI is constantly, by definition, evolving. It's only as good as its code and it's only as good as the data and the learning that the computer has gone through. I think you need to pick a partner, that if they themselves don't have a long track record, that they partnered themselves with somebody that does.

You want them to continue to have something approaching a cutting-edge offering. We've really only begun scratching the surface of this. In comparison, the tools you're buying now are an iPhone 1, not the iPhone 8.

Moving away from AI a bit, the General Data Protection Regulation deadline is almost upon us. Have you had to work on compliance?

GDPR is certainly front and center. We do not have any legal operations in Europe. The legislation is written to be fairly broad, which doesn't necessarily preclude us from those compliance requirements.

But we are definitely focused on that. We have significant operations in Canada, which has different data privacy regulations than the U.S. that are closer to the old EU standard.

As a data company, at some level, we have a significant amount of client data. A lot of these requirements, if you look at the New York cyber regulations, we were already doing most of what they required anyway. There were a handful of things that we had to tweak. With or without the regulations you should be doing those anyway. It's just good business.