Commercial insurance underwriters fear losing senior expertise more than being replaced by artificial intelligence (AI), yet firms are investing least in preserving that knowledge, according to a recent hyperexponential (hx) survey.
The hx Underwriting Edge Survey was conducted by independent research firm Coleman Parkes on behalf of hyperexponential.
It surveyed 350 Chief Underwriting Officers, Heads of Underwriting / Line of Business, and Senior and Lead Underwriters working in commercial and specialty Property & Casualty insurance, split 43% UK and 57% US.
It found that 44% of underwriters believe that senior judgment leaving the industry, without being passed on, is their top fear, with coaching and knowledge-transfer technology receiving only 8% of firms’ investment over the next 12 to 18 months, the lowest of any category surveyed.
Only 15% of the respondents said their firm has found a way to capture what its best underwriters know. Two in five respondents admitted that critical judgement exists either poorly documented or strictly inside a colleague’s head.
Notably, worry that AI will replace the underwriter’s job entirely keeps falling. Only 5% of respondents called it an urgent issue, compared to 18% in 2024, and 11% in 2025.
Asked what separates a good underwriter from an excellent one, three things top the list: risk selection and appetite discipline (48%), technical depth on complex risks (43%), and commercial instinct on price (38%).
According to the hyperexponential Underwriting Edge Report, those closest to the work remain the most skeptical about AI. Although 51% of underwriters cite manual admin time savings as AI’s main benefit, only 21% report improved decision quality.
Furthermore, just 12% of senior and lead underwriters feel AI is fully integrated into their workflow, compared to 29% of Chief Underwriting Officers, showing confidence grows further from the actual work.
That skepticism is not a rejection of AI, hx noted, it is a redirection. Forced to choose, underwriters picked richer context over raw speed.
73% preferred real-time alerts on portfolio drift and exposure over auto-refresh reports, and 70% prioritise transparent pricing models that surface prior risks over raw processing speed.
While AI reduces manual admin from 17% to an ideal 8% of the underwriting week—freeing time for brokers, portfolio reviews, and coaching—it fails to bridge the judgment gap.
The report also found some key obstacles to decision-making, which include inconsistent data (44%), rushing to bind over analysing (38%), and insufficient context on prior risks (35%). Ultimately, the issue is bad inputs and limited time for judgment, not underwriter skill.
“AI suggests, I approve” was the top-ranked way underwriters want AI to engage with every task tested. The more judgment a task demands, the less underwriters trust AI to handle it alone: asked how they’d want AI to handle each part of the job, from acting fully on its own to staying out of it entirely, the share comfortable with full AI autonomy falls from 34% on data ingestion to just 8% on coaching junior underwriters.
“Underwriters have told us exactly what they’re afraid of losing, and exactly what they want protected,” said Jamie Wilson, VP of Strategy at hyperexponential. “The industry keeps building AI to move faster, and carriers keep funding it that way. What this survey shows is that speed was never the scarce resource. Judgment is, and right now it’s leaving faster than anyone is capturing it, or investing in catching it.”
