In theory, technology should be making commercial insurance cleaner, faster and safer. In practice, however, it may be quietly creating some of the most serious exposures brokers have faced in years.
Ask Graeme Finnell (pictured) which commercial risks are the most challenging to place right now and he doesn’t hesitate. From his vantage point as vice president, casualty / broker relations at ABEX, it starts with property coverage – and it doesn’t stop there.
“As a P&C market within the MGA space, property-driven weather and CAT losses continue to be a real concern and challenge from an underwriting point of view,” he says. Flood, quake, fire and wind – across Alberta, Manitoba, Ontario and the Atlantic provinces – all make underwriting risk in those areas very challenging. Good underwriting discipline and mapping tools are critical, he adds, to avoid becoming overexposed to these kinds of losses.
Read more: Are portals helping or hindering MGAs?
Then there’s the other side of “hard to place”: specialty liability.
Even as markets have learned to talk about cyber and professional liability more fluently, Finnell doesn’t see the concern easing. As he frames it, every business owner should be asking: do I have the right coverage, and do I have enough limit? A single cyber event that locks a business out of its systems or exposes critical data can spiral into existential damage – not just business interruption and reputational harm, but also shareholder anger and potential D&O fallout.
“Imagine the impact to a shareholder from a cyber‑related shutdown of a business that can’t continue to run or make profit,” he says. “That’s a huge exposure.”
All of that would be challenging enough in a traditional, conversation‑driven marketplace. But the industry is increasingly routing small and mid‑sized risks through portals and automated platforms, often with AI working behind the scenes.
From a workflow perspective, it’s easy to see the attraction. “A lot of business is going to portals,” Finnell notes. “The advancement of AI makes it quick and easy.” For brokers under pressure to quote and bind rapidly, the promise of speed and standardization is seductive.
His concern is what gets lost.
Portals, by design, standardize risk through predefined questions and outputs. They thrive on neat categories and tidy yes/no answers. Small businesses, by definition, don’t always fit within that model.
“Portals are meant to standardize risk,” he says. “But a lot of small businesses have little nuances that don’t translate well into predefined environments.” Those nuances – a sideline operation, an unusual subcontracting arrangement, a quirky set of premises – are precisely where coverage gaps and claims disputes tend to emerge.
Finnell is watching a trend that should make brokers uncomfortable: plaintiff lawyers increasingly focusing on broker conduct when claims go wrong, rather than solely on carriers. When coverage is denied, the key question is shifting from “what did the insurer do?” to “what did the broker fail to see, ask or explain?”
“When claims don’t get paid, they’re going after brokers who’ve not done their homework or due diligence,” he says. The common thread in many of those cases, he observes, is over-reliance on portals for transactional speed instead of deeper risk discovery.
For Finnell, the heart of brokerage hasn’t changed. The job is still risk management and solution‑finding with the client. The differences between one small business and another – even in the same class – often surface “through conversations, not checklists.”
That’s where he sees real danger in a purely tech‑driven model. A portal can’t sit across from a client and hear the hesitation in an owner’s voice. It can’t ask the follow‑up question that reveals a new exposure. It can’t challenge a client who insists “we don’t really do that” when, in fact, they do.
Technology isn’t going away, nor should it. But in a world of more severe weather events, sharper cyber incidents and more aggressive plaintiff bars, Finnell’s concern is clear: brokers who let portals and AI do the thinking for them may find out, too late, that the real risk they underwrote was their own.
