Artificial intelligence has moved from novelty to necessity. Once confined to tech labs and futurist conferences, it is now quietly threading its way into underwriting systems, claims platforms and customer contact centres across the UK insurance market.
For an industry long defined by documentation, analysis and regulation, AI represents both promise and peril — an innovation that could raise productivity even as it threatens to hollow out the very roles that keep the sector running.
A major study published in Harvard Business Review by Evercore ISI and venture studio Visionary Future analysed more than 160 million jobs to gauge how generative AI might reshape global workforces. The conclusion was clear: nearly every occupation carries some level of exposure to automation.
While that may sound abstract, its implications for British insurers are concrete. Many of the tasks most vulnerable to AI are central to the business of insurance — the clerical and analytical work that underpins underwriting, claims handling and compliance.
As the report notes, “AI will emerge not merely as a technological marvel, but as a beacon of hope in addressing demographic and productivity challenges.” Yet for some, that beacon may illuminate a shrinking career path.
Historically, automation threatened manual labour first — factory lines, logistics, production. Generative AI reverses that logic. Its strength lies not in physical efficiency but in linguistic precision: summarising, comparing, calculating, drafting and deciding.
According to the Evercore study, cognitive abilities such as information ordering and memorisation are highly exposed to AI replication. The more repetitive and rules-based a role, the greater the vulnerability. Creative, interpersonal or judgment-based tasks, by contrast, remain largely resistant.
In other words, it is not the actuary on the trading floor who faces the greatest threat, but the administrator in the back office.
Drawing on findings from Evercore ISI, the OECD, the IMF and the International Labour Organization, the following roles stand at the highest risk of disruption or redesign in the coming decade:
These positions share a common thread: they depend heavily on structured information, templates and rules — precisely where large-language models excel.
In insurance, the exposure extends across the value chain: policy administration, claims triage, fraud detection and document preparation. The very processes that make the industry methodical and precise now make it susceptible to automation.
AI’s immediate appeal to insurers is obvious. Productivity growth in the UK has stagnated since the financial crisis, while cost pressures from inflation and regulatory oversight continue to climb.
A claims platform that can draft correspondence automatically or a chatbot capable of answering 60 per cent of policyholder queries promises enormous savings. Yet such efficiency carries a cost. The entry-level administrative and support roles being replaced are also the sector’s training ground — the foundation on which technical and managerial expertise is built.
If AI strips away these formative experiences, the industry could find itself efficient but shallow: rich in data, poor in development.
AI is also encroaching on higher-paid work. Financial analysts, solicitors, accountants and underwriters now rely on AI tools that summarise documents, compare contracts and calculate probabilities at unprecedented speed.
Evercore’s analysis shows that roles earning more than £80,000 a year have some of the highest AI exposure, but they also stand to gain the most from augmentation. Those who learn to use AI intelligently will thrive; those who ignore it may find their expertise diminished by the very systems they once commissioned.
Already, several Lloyd’s managing agents are experimenting with AI to assist in risk evaluation and documentation. Brokers are trialling tools to draft placement summaries. Claims teams use algorithms to flag anomalies before investigators ever read a file.
The technology is not replacing people yet — but it is changing what those people do.
Insurance remains a business built on trust and compliance. Regulators will not allow underwriting or claims decisions to be delegated wholly to machines, and policyholders expect human judgment when outcomes are contested.
AI may accelerate processes, but responsibility cannot be automated. Insurers must therefore tread carefully: they must integrate new systems without compromising accountability or fairness. The Financial Conduct Authority has already indicated that governance, transparency and data ethics will form part of its scrutiny of AI adoption across financial services.
For business leaders, the question is not whether AI will reshape insurance work, but how to manage the transition. That means conducting workforce exposure assessments, retraining administrative staff for analytical or client-facing roles, and redesigning job families around human strengths — empathy, reasoning and ethical decision-making.
Evercore’s data suggest that roughly a third of all job functions can already be enhanced by AI tools. The task for leadership is to ensure that the other two-thirds — the human judgement, experience and negotiation — are elevated, not eroded.
The next decade will not spell the end of white-collar employment, but it will demand a redefinition of value. Insurance has always prized accuracy and efficiency; soon it will prize discernment and trust just as highly.
As algorithms take over the arithmetic, people will be judged on their ability to interpret, explain and connect — to understand not just what the model predicts, but why it matters.
For Britain’s insurers, success will not hinge on how quickly they adopt AI, but on how wisely they keep humanity in the loop.
AI is coming for the forms, not the relationships. And in the halls of the City and the underwriting floors of Lime Street, the firms that survive will be those that see technology not as a substitute for people, but as a reason to make people matter more.
