As more of that journey is delegated, insurers face a serious question. Will they remain the trusted guide for retirement income, protection and wealth transfer? Or will AI agents become the custodian of customer relationships?
UK life and retirement insurers are being pulled in two directions. Between 50 and 60 percent of operating expenditure is still absorbed by policy administration on decades-old systems. At the same time, the industry must close a £5 trillion savings gap and £2 trillion protection gap, while supporting millions making complex decumulation decisions without advice.
Agentic AI can address both pressures: taking structural cost out of legacy operations while enabling propositions that are affordable, relevant and engaging at scale. A customer approaching retirement, for example, could use an agent to analyse pension pots, model income options and prompt timely action, making support once limited to the affluent viable for the mass market.
This shift is happening as Consumer Duty requirements tighten, the population ages, and a £5.5 trillion intergenerational wealth transfer accelerates. With only 9 percent of consumers receiving regulated advice, AI agents, workplace platforms and aggregators are stepping in to shape discovery and comparison.
For thirty years, marketing teams asked Insurers whether a consumer or broker could find them. The next question is whether the consumer’s AI agent can. Insurers whose products are not machine-readable, easy to explain or simple to integrate risk losing the moments that matter, providing the regulated product but losing the relationship, data and influence.
As agentic AI reshapes how cover is discovered, advised on and bought, insurers need to decide whether to be an orchestrator, partner or component supplier. Those who hedge across all three will struggle to make consistent choices on data, technology and distribution. Those who commit can align investment with strategy and defend their position.
None of this works without strong foundations: clean, connected data, AI-ready governance built in from the start, and a clear division of labour between advisers, AI copilots and autonomous agents.
Agentic AI will change more than efficiency. It will reshape who guides the customer, who is trusted with the decision, and who owns the industry’s most valuable relationship.