Briefing: The price of expertise in professional services

A recent crisis in the legal industry looks like a cautionary tale for the service industry’s use of AI.

Top US law firm Sullivan and Cromwell (S&C) reportedly faced embarrassment, when Boies Schiller Flexner, the lawyers acting for the other side, noted a submission to a bankruptcy court judge contained so-called AI hallucinations – including the citation of a case that did not exist.

The legal industry’s most recent AI embarrassment may actually expose a deeper weakness in insurance – we still struggle to price expertise itself. 

S&C partners can reportedly command over $2,000 per hour, although one suspects it was someone a little lower down the food chain who let the firm’s editorial standards slip.

Whether the episode has shaken trust somewhat for the firm or the industry remains to be seen, but it is relevant because the real contrast is not actually AI-risk itself, but how each profession prices accountability.

Beyond highlighting AI’s propensity to distort risk across already problematic professional indemnity classes, the episode exposes something stranger about insurance itself. It might serve to remind us of how bizarre the insurance expertise pricing model is in comparison.

When one selects a law firm, one may choose the seniority of those they wish to undertake the work – and the hourly rate reflects that.

There is surely an implicit understanding that experience is priced accordingly and must surely affect propensity to kick up a stink if said legal representation has fallen foul of the AI bubble, as tempting as its seductive, time-saving tools may be.

In this example, S&C’s entire reputation, and therefore all its equity partners, likely ended up in the crosshairs, despite the fact the organisation has robust AI processes and even advises OpenAI on ethics.

Is such a reputational outcome compatible with the pricing model most law firms have developed, one where you quite literally pay for what you get? 

Broker expertise

The contrast between the legal expertise model with the insurance broker model is striking.

I couldn’t find it, but I was sure I’d written something on this subject before – commission tiers. Suspend disbelief and imagine a brokerage where the seniority of those giving advice or placement support was also tiered by expertise and accountability. Where the brokerage commission is somehow reflective of the various layers of professional expertise that have helped produce it.

Junior brokers charge less than the premium senior advisory staff and the same is true for much larger, enterprise or reinsurance risks – reflecting the value and accountability on offer. 

Broker PI can be expensive these days. Would buyers perceive professional responsibility differently if expertise itself were transparently tiered? A client is unlikely to explicitly factor in the specific experience of their representative at a broker – although that is surely an implicit factor.

In Stephen Catlin’s still relevant 2017 industry biography, Risk and Reward, the chapter on brokers remains fascinating.

“We need each other, but can we get along?” he begins. He rightly points out that the broker works for the buyer of insurance and, more often than not, is unlikely to be remunerated by fees for time as other professional services are.

Whilst much larger programmes are utilising fees, the commission, particularly across supply chains from retail to wholesale, is a fascinating dynamic. 

The commission model

The pressure of the commission model in insurance is clear – and materially affects the buyer’s perception of the entire product.

Commission transparency aside, every time a brokerage takes a blended commission, it is effectively spreading the risk across the entire firm.

A law firm doesn’t really do that, as there’s far more internal accountability between partners if the firm’s insurance or reputation has to be drawn down. One wonders what the consequences will be for the partners at the S&C team in the case above!

Follow the rabbit hole further on commission, though. Consider whether insurers are even correct to offer brokers quotes for free. Many in the industry would acknowledge that for every 100 brokers capacity will trade with, 20% likely bring 80% of the support.

It’s provocative to think that the entire downstream pricing model could be different, if quotes cost money. Or if agencies carried minimum premium commitments. That observation may make a few shoulders shudder, but it could also create a new asset class – such as annual capacity options following bids and auctions.

In a parallel insurance market, the broker that’s chartered or qualified is able to charge more commission to the clients because the insurer enables it. Quotes for clients are chargeable because capacity itself is. If expertise has value, then arguably capacity must too. 

Catlin talks to this himself, regaling readers with a brilliant anecdote concerning a broker where he concludes a conversation about higher commissions by asking if the broker sees them offering a good service. He’s told they are a preferred carrier.

“When have you ever paid me for that service, just so I understand?” he retorts. 

It’s been almost a decade since that book. Despite the so-called wave of disruption, and despite the proliferation of many more MGAs, no different model has emerged. 

After all, it would take a brave MGA or insurer to move first – or perhaps that’s a new blueprint for an insurance market? 

This all ties rather neatly to perceived value. As our industry grapples with the seismic nature and arrival of AI, unlike our legal cousins we have yet to even crack the value-added perception from insurer to broker to buyer.

It’s hard not to think it’s a can we can only kick down the road for so long. 

Such an experiment is unlikely in the near term. Commission, insurer to broker negotiations and complex explanations for remuneration will continue.

For, as Stephen Catlin noted a decade ago in his anecdote, “that’s the way the system works and it would be difficult, if not impossible, to change it”. 

AI may not disrupt insurance first through underwriting or claims. It may instead force the market to finally confront a much older question – what is risk expertise actually worth?