This piece is by Jenny Burns, CEO, Magnetic
The first recorded insurance policy dates back to 1347. For nearly seven centuries, commercial insurance has excelled at one thing above all else: pricing what might happen. But AI is about to transform that accumulated knowledge overnight.
The shift is already underway. The world is starting to move from reactive assessment to AI-fuelled predictive, and in some cases preventive, risk management. Because of this, the challenge for insurance leaders is not whether they should be adopting AI, but where they should be drawing the line. Because AI is not a free lunch. Like everything in life there are trade-offs. Yes, it can accelerate decision-making and scale operations, but without thoughtful role design, learning pathways and decision governance, your company’s hard-won expertise could be eroded faster than you gain new value.
Contrary to a lot of noise in the industry, your success doesn’t lie purely in the speed of AI adoption, it lies in understanding where humans must remain at the centre of the process.
Human judgement remains your strategic advantage
Almost every insurer is experimenting with AI in some form – underwriting, claims, reinsurance, fraud detection. Generative models and predictive analytics excel at spotting patterns, detecting anomalies and processing documents. The efficiency gains are real.
Aviva deployed 80 AI models across claims, cutting liability assessment times by 23 days, improving routing accuracy by 30% and reducing complaints by 65%. This saved over £60m across the process.
But those gains come with pressure. Automating routine work can erode the human judgement that makes the hard calls possible in the first place while simultaneously undercutting the next generation, with junior underwriters and claims handlers getting fewer opportunities to develop the instincts needed for complex or unusual cases.
It’s no surprise then that over 40% of entry-level employees in the insurance industry believe that technological change will impact their jobs to a major extent over the next three years. This is compounded by a third of those same employees saying they’re worried about how AI will affect their careers.
As Rob Flynn, ex-UK Commercial Lines Chief Transformation Director at Intact, puts it: “There’s always a pull to roll out AI fast, but we can’t lose sight of getting it right for our customers. AI can make things more efficient, but human judgement is still what makes the tricky calls work”.
Complex claims, emerging risks, and high-value policies all demand nuance, empathy and cultural understanding that AI can’t replicate. The human capability required here is a genuine competitive advantage. But only if it is built deliberately into the workflow.
With AI absorbing routine tasks, insurers must redesign roles, career paths and learning programmes. The risk of not doing so? Becoming very efficient at the predictable but dangerously exposed when the unexpected hits.
Trust is the aim of the game
Insurers using predictive analytics have seen loss ratios fall by up to 80%. The future of the industry is moving from reactive payout to proactive risk management. In that sense, the direction of travel is clear.
But at the heart of it, prevention and claims excellence share the same principle: AI amplifies human capability. Trust and judgement remain at the centre of every decision. Both traits that require human empathy and understanding.
That trust is earned when customers know they can rely on insurers to restore certainty in their lives. Every customer, whatever their claim and whatever their circumstances, should experience the same clarity, fairness and quality of decision, whether they’re dealing with a skilled handler or an AI-enabled digital journey.
Automation must therefore uphold the standards of judgement and professionalism that define the best human teams, with clear accountability for every outcome. Because AI can mitigate risk but it can also create it – algorithmic failure, data liabilities, environmental consequences and systemic disruptions are all worst case possibilities. Who prices those? Automate existing processes blindly and you risk blind spots with potentially significant consequences.
AI is forcing insurance leaders to make hard decisions about employees, purpose and their relationships with customers and society. To ensure the decisions that matter stay in human hands and companies retain the faith of their customers, three things are non-negotiable.
First, we must create systems that keep humans in the loop and build and maintain trust with teams and customers alike. Second, we must retain learning and career paths designed to maintain edge case expertise. We can’t expect the next generation to learn the key skills of insurance if we do not give them the opportunities to do so. Third, we need people first integration that redesigns roles and decision rights, not just retrofits AI onto existing structures.
Get all three right and AI becomes what it should be: a tool that makes insurers more efficient and more expert. One that earns, and keeps, the trust of the customers and businesses that depend on them.
Be the first to comment