Iprism took to the airwaves this month for an Insurance Post webinar, ‘AI in Action – Underwriting Lessons from the Front Line.’ During this session, our Operations Manager, James Walker joined an expert panel to discuss the integration of AI tools in underwriting systems. The conversation explored the problems AI is already solving, the benefits the sector is beginning to see, and importantly, the potential barriers: from regulation, and workplace culture, to the importance of human judgment.
Keeping the human in the loop
Despite the increasingly sophisticated capabilities of AI, James was clear that having a human influence is absolutely essential. When asked about Iprism’s integration of AI, he emphasised that to keep our company values, AI must remain “an assistant, not a decision maker in and of itself.”
Our team already uses AI for terminology checks, clarification and research, and early-stage analysis, but he reiterated that no AI system currently in place acts as an independent underwriting engine. And crucially, he doesn’t believe it should.
James explained the reasons behind our more cautious approach, and our desire to innovate sustainably, with regulation. As he shared with the panel, responsible adoption requires clarity on compliance, governance and data protection: “There hasn’t been a burning need or a single part of the business where we’ve said: we must automate this now using AI”. The business already operates highly effectively with the current team of technical underwriting experts, and automating bespoke underwriting processes would be a huge change. “We’re acutely aware that trust is needed, from underwriters, from compliance, from our insurer partners, before any major step forward.”
Elevation not elimination
One of James’ key contributions was his recognition of what may be at stake. AI might improve efficiencies, but without careful implementation, it could cause insecurity in staff at all levels. Senior underwriters may fear being sidelined; junior staff may feel inadequate if traditional entry-level tasks are automated. James was clear that the industry can’t allow this to happen. “It’s really important that new starters see this as an opportunity to accelerate their growth, rather than a replacement.”
At Iprism, we see AI as a tool that helps underwriting assistants build technical confidence faster, prepare better referrals, and speak the language of underwriting earlier in their careers. “Hopefully it will accelerate their knowledge and expertise beyond years of experience.” Meaning AI won’t eliminate entry-level roles; instead, it will elevate them. Assistants can speed through the learning period of looking through workflows and waiting for sign-offs, and spend more time engaging in deeper analysis. AI can kickstart their careers and allow assistants to participate more meaningfully in conversations with brokers and develop a stronger understanding of risk.
Understanding risk is essential to the underwriting process, so while underwriters may be tempted to trust outputs unquestioningly, James reminded the panel that expert human judgment remains the heart of underwriting. AI can inform our thinking, but it cannot replace professional analysis.
AI could also add real value in internal referrals from junior to senior underwriters. Rather than making decisions, AI could help underwriting assistants prepare fuller, more informed submissions to senior underwriters, suggesting missing information, highlighting potential issues. Conversations could begin at a higher level, speeding up broker responses but also deepening the development of juniors through more valuable conversations with senior staff. AI then becomes a bridge that narrows the experience gap, nurturing professional development.
A future of integration, not replacement
At Iprism, our underwriting reputation comes from more than just our efficiency. It’s our team of experts, their years of technical underwriting expertise, deep knowledge of their lines of business and their ability to provide effortless underwriting for our partners and brokers. Though we’re excited to see how AI becomes woven into underwriting processes, it’s important to remember that in a market still driven on human interaction, AI should never replace the very real people responsible for risk. “This is never going to be an underwriter in a box […] This is a tool that’s going to help individuals progress and develop their understanding at rates we’ve never seen.”
Share
