While much of the insurance industry is contending with unprecedented volatility, purchasers of professional lines have enjoyed not just stable but even declining rates.  

This varies by line, but on the whole, “capacity between carriers for private D&O placements remains strong,” according to David Garrigus, EVP, management and professional liability practice leader, HUB International. “Underwriters are aggressively cutting premiums, giving automatic renewals and two-year policies to maintain business. D&O insurance for public companies remains highly competitive. There is plenty of capacity, and underwriters remain aggressive. Additionally, E&O rates for consultants, real estate and trustees remain competitive as underwriters compete for their business.” 

“It remains a very competitive marketplace. There’s ample capacity for sure,” agreed Stephen Swartley, CUO, executive assurance & SVP, growth & middle market, Arch Insurance. “Towards the end of 2022, more capacity came in and rate came out of the business. We’re still in that part of the cycle now, although it’s starting to be more flat; generally, accounts are still seeing decreases, but less so than earlier in the year, and certainly less so than last year.” 

But while rates may still be trending down, insureds need to keep their guard — and their limits — up; triggers for unexpected large losses abound, especially as class actions and nuclear verdicts signal increased volatility ahead. 

“I’m speaking specifically to D&O, but it’s across the board,” said Keith Riccio, EVP, commercial public and private liability, Nationwide. “Litigation costs are higher, and average settlements have gone up. The numbers that are being kicked around — something that 10 years ago might have been $50 million, now it’s $100 million. Market caps are higher now, so that number has naturally gone up.” 

Securities class actions, Swartley noted, “have seen an uptick. That was up 12% last year; it’s up another 5% so far this year … and I think most people expect that uptick to continue through the rest of this year, and some would say to persist into next year.” 

Whatever drama lies ahead, it will feature a new starring character: artificial intelligence. 

A New Variable in the Mix

AI holds enormous potential to improve efficiencies in virtually all industries, but unlike technologies of the past — passive tools to be used — AI is more like an active agent, making it all the more likely to trigger professional liability losses, and nowhere is this potential more apparent than in employment practices.  

Stephen Swartley, CUO, executive assurance & SVP, growth & middle market, Arch Insurance

“There’s a bigger conversation around AI and the ethical and professional implementation of AI,” said Keith Savino, practice leader for professional and cyber liability, Trucordia (formerly PCF Insurance). “The NAIC put out a paper on this about four years ago, and other states and industry groups are coming out with recommendations and positions around its implementation, particularly generative AI … On the surface, [using AI to make hiring decisions] sounds wonderful. But in practice, we just don’t know yet. It’s new.” 

New, but by no means rare. In January 2023, the EEOC cited estimates that “as many as 83% of employers and up to 99% of Fortune 500 companies now use some form of automated tool to screen or rank candidates for hire.” 

“AI is a large issue for the industry as a whole. You could potentially see liability there under an EPLI policy or a D&O policy, depending on how it’s worded, if there were any exclusions carved back, things of that nature,” said Riccio. “The main issue around AI — and it’s not just AI; it’s any event-driven type of coverage, or any exposure in general — is whether your coverage is adequate for your business needs.” 

Particularly if they’re using AI in personnel management, employers should have an EPLI policy in place. “This would fall within the EPL space in terms of discrimination, both in targeting applicants and in moving people along within the interview process,” explained Jeremy Moen, SVP, professional liability claims, Arch Insurance, before noting that “not everybody buys EPL.” 

“Most companies aren’t going to discriminate on purpose, so the risk of the AI model is that they’ll feed it their own hiring or promotion data, and that will reflect back some inherent discriminatory practices that they had and maybe didn’t even know about,” Swartley said. “All the AI will do is continue that pattern, and if that is indeed the case, they could get sued.” 

While this would likely be covered under a GPL policy, Swartley said, “you have to have enough limit to contemplate something like that. It’s not the sort of claim that you would necessarily jot out on your board of most likely claims, but it’s one that could happen.” 

“A well-structured EPL program absolutely ought to pick up discrimination, whether it’s AI-derived or otherwise, but be very careful around the exclusions that you have in your policy that insurers may be trying to introduce into the contract,” noted Rik Goyton, SVP at CAC Specialty. 

“In all of my recent employment practice liability submissions with underwriting, AI is now certainly a topic of interest, where underwriters are asking whether or not companies are using AI to scan résumés and chatbots to interview candidates,” Goyton added. “They’re definitely focused on it.” 

When Problems Snowball 

Because AI is used more systematically than a single hiring manager, it has the potential to wreak mismanagement company-wide — widely enough, perhaps, to trigger a D&O policy. 

“It’s absolutely a possibility,” said Goyton. “It would typically have to rise to a quantum of loss that would become interesting to a plaintiffs’ bar to want to take on such a case, but … we’ve seen cases where companies had an awareness around glitchy software and didn’t necessarily address it properly.” 

Goyton continued, “What does the board know, what does the board do or not do about it, and what is the board management saying to the street about its risk and exposure? I have not seen that type of employment [event] rise to a D&O claim in my own portfolio yet, but I could definitely see the potential.” 

In fact, when it comes to what a board of directors is “saying to the street” — not only when it comes to its potential liabilities in employment practices, but in anything tied to its use of artificial intelligence and its potential for failure, or even its potential for success — securities class actions are a possibility. Far beyond the scope of using AI to scan résumés, companies caught up in the hype cycle of AI are making public disclosures about their AI practices and how it might boost their future gains. 

“For publicly traded companies, this is absolutely in the wheelhouse of what public company D&O is designed to cover,” Goyton said, noting that “we’ve seen the words ‘AI washing’ come up from the SEC recently, and they’re starting to look at AI-related disclosure issues with various small companies. You have a product, you claim it’s AI, you indicate that your AI is capable of doing certain things, and it delivers or it doesn’t. A lot of those issues come back in the concept of disclosure.”

Keith Riccio, EVP, commercial public and private liability, Nationwide

Class action lawsuits triggered by the (mis)use of AI are no hypothetical. “AI-related securities class action is a trend that we have been seeing for a couple years now,” Swartley said, “but they’re cases that have something to do with the company embellishing their ability to make money or make profits related to AI, given how hot that is now, and then not actually being able to deliver — we’ve seen that as one class action trend.” 

“You could envision a scenario where the leading HR software becomes subjected to claims of discrimination, and that has ramifications through a much broader marketplace,” Moen added. 

“It’s ripe for a class-action type claim, and those are going to be more expensive than single-party claims,” Swartley said. “I put it in the emerging litigation area. As we sit here today, there haven’t been, to my knowledge, many claims like we’ve described. It’s one of those things that you have to contemplate as you’re buying your limits going forward. It’s something that hasn’t hit many companies yet, but might start to become a real issue.” 

Cover Your Assets 

While the triggers may be novel, the liabilities themselves are not, and existing policies may be up to the challenge of securing their risks. But insureds and their brokers should examine their policy language closely — and, just as importantly, understand how quickly the costs of an event can spiral out of control.  

“A lot of times, insureds are underinsured,” said Riccio. “They feel like they have a package of insurance that meets all their needs, and they don’t really view insurance as an asset — they view it as a cost. They just want to get the coverage they need to satisfy whatever requirements they have for the cheapest price.” 

But as their use of AI grows, so do their potential exposures, insureds and their brokers should “look at industry benchmarks with regards to D&O, E&O, cyber and professional liability. Your broker should be providing you with information about companies of your like size,” Riccio said. “If your peers are buying $50 million or $100 million in D&O, $50 million in cyber, $50 million in miscellaneous professional liability — yeah, it’s going to be costly. And it might seem expensive, but you know what’s also expensive? Having a claim and not having coverage, and you’re on the hook for that. It could potentially put a small company out of business, or give a large commercial account a big hit to their balance sheet. The industry needs to make sure that they’re buying the right coverage — and enough of it.” 

But, Riccio warns, no limit is high enough if its coverage is not the right fit. “Don’t trade price for coverage. Coverage should be the most important thing about your insurance buy, even before the limit. You could buy $200 million, but if you don’t have a good insurance program there — if the underwriter threw all these exclusions on — then what are we buying?” &