Last week, the Portage team traveled to Bermuda for a series of conversations with leaders across the island's government, business, and (re)insurance communities. Two takeaways stood out: Bermuda's appetite for practical, well-governed AI runs far wider than its flagship (re)insurance industry, and that appetite is grounded in a clear-eyed view of what the technology can and can't do.
We had those conversations alongside Dr. Alok Aggarwal, founder of Scry AI and the author of The Fourth Industrial Revolution and 100 Years of AI. Scry AI is a close partner of ours, and Alok has spent years researching how this technology will reshape economies and labor markets. His perspective sharpened every discussion.
A wider aperture than we anticipated
The leaders we met were curious about AI's role well beyond the risk industry — in agriculture, where a small island has to think hard about food security and productivity; in education, where preparing the next generation is inseparable from preparing the workforce; and in healthcare, where an ageing population is reshaping demand faster than almost anywhere.
That range mirrors a point Alok makes often: AI isn't a single story about machines replacing people. It's a set of overlapping shifts happening at once. His research projects roughly 400 million jobs' worth of work could be automated by 2050 — but also that ageing populations, climate adaptation, infrastructure, and entirely new industries will create even more work than they displace. The result he foresees isn't mass unemployment; it's a deepening labor shortage, and a premium on people who use AI well and bring genuine subject-matter expertise.
His caution resonated: move deliberately. Some companies have moved too fast, cut too deep, and had to rehire much of what they let go — because they underestimated how much institutional knowledge and human judgment their operations actually run on. For an industry built on expertise like (re)insurance, that lesson lands hard.
Not whether to adopt AI, but how?
Across nearly every conversation, the through-line was the same: not whether to adopt AI, but how to do it in a way that produces measurable results and holds up to scrutiny.
That's exactly the question we like to be asked.
Where Portage sees the real payoff
The (re)insurance industry, in many ways, an ideal proving ground for AI. So much complex, high-value work is still done manually — spreadsheets, re-keyed data, and painstaking submission and bordereaux management. There's meaningful efficiency to be gained in open-market and delegated-business pricing workflows, and those gains flow straight to the expense ratio.
But the bigger prize is the data itself. Insurers sit on enormous volumes of it, and AI's real power is turning that data into actionable insight — pricing signals, risk patterns, and decisions that were previously buried in unstructured files.
We're also clear-eyed about the industry's skepticism, and we think it's earned. Too many projects stall at proof-of-concept and never touch the P&L. That's why we adopted a value-based rather than usage-based model: we don't ask clients to pay until they see real results. We'll do the work on our dime, and if we don't deliver value, we go away. It's a simple way to align our incentives with our clients' — and to cut through the hype with something tangible.
Read more
Our thanks to The Royal Gazette's Business Editor Jonathan Kent, who sat down with the team to talk through what AI means for the island's economy. You can read his write-up here: We have time to adapt to AI's huge job impact, says expert.