The insurance industry has spent three years asking whether artificial intelligence will take its jobs. PwC's 2026 Global AI Jobs Barometer, published yesterday and based on analysis of more than a billion job advertisements across 27 countries, suggests that is the wrong question. The more important one - and the one with real consequences for every person working in insurance - is which jobs AI will make more valuable, and which it will quietly hollow out.
The answer turns on a distinction PwC calls professionalisation versus democratisation. In professionalised roles, AI strips away routine tasks and leaves the more complex, expert work to people, making those roles harder, better paid and more in demand. In democratised roles, AI absorbs the expert tasks instead, leaving the less demanding work behind and gradually reducing the premium attached to those jobs. The report's data shows professionalised roles growing twice as fast in headcount and seeing 42% faster wage growth since 2021. Democratised roles are growing, but more slowly, and their wages are falling behind.
Insurance sits directly in the path of this divergence. Credit and loans officers and finance managers appear in the report's worked examples of democratised roles - positions where AI is already performing tasks that previously required specialist expertise. The implication for insurance is pointed: large swathes of routine claims handling, policy processing and standard underwriting assessment belong in the same category.
PwC's sector-level data produces a finding that deserves more attention than it is likely to receive. The wage premium for AI skills in financial services - the gap between what employers pay for roles requiring AI capabilities versus comparable roles that do not - stands at 53%. That is a significant number in absolute terms. But it is also nine percentage points below the cross-sector average of 62%, and well below consumer markets at 118% or technology, media and telecoms at 84%.
The gap suggests financial services, including insurance, has been slower than other sectors to price AI skills into its pay structures - which in a tightening talent market for AI-fluent professionals is a recruiting vulnerability. The UK insurance industry recorded a 70% increase in job listings requiring AI expertise over the past year, but the compensation structures around those roles may not yet reflect what the broader labour market is willing to pay for those skills.
PWC BAROMETER: KEY FIGURES FOR INSURANCE
The professionalisation dynamic is already visible in insurance hiring. Insurance underwriters are shifting toward reviewing AI outputs, managing exceptions, and applying judgement in complex cases as AI takes on first-pass risk assessment. The role is not disappearing. It is becoming more senior, faster, and harder to fill with candidates who have the right blend of technical fluency and underwriting judgement. PwC's separate analysis of financial services firms found that insurance companies exploring AI are discovering it works best when it supports expert decision-making rather than replacing it - a finding that reinforces rather than challenges the professionalisation thesis.
The graduate pipeline sits on the other side of that equation. PwC found that entry-level vacancies in the most AI-exposed roles have flatlined globally - the only quartile where that is the case. Among those AI-exposed junior roles that are still growing, the common thread is that they have been effectively seniorised: now requiring leadership, stakeholder management and strategic thinking that previously took years of experience to develop. Entry-level roles demanding ten or more of these senior capabilities are growing at 35%; those that have not evolved have fallen 10%. For an industry facing the retirement of roughly half its workforce within 15 years, that compression of the entry-level pipeline is a structural problem as much as a skills one.
The report's most commercially pointed finding concerns what PwC calls the superstar effect. Among the most AI-exposed companies, the top 20% have achieved productivity growth of 163% since 2018, compared with 33.5% for the most AI-exposed group as a whole - a gap that has tripled since 2022. The differentiator is not how much AI a company deploys but what it deploys it for: the superstars are using it to pursue growth and create new forms of value, not simply to reduce headcount or automate existing processes. Headcount at those companies is growing 45% faster than at their less AI-exposed peers, and wages are rising faster too.
For the more than half of UK insurers who now have AI embedded in core functions but are struggling to translate isolated deployments into enterprise-wide results, this is the crux of the challenge. PwC's CEO survey found only 8% of chief executives report AI generating more than a slight increase in revenue so far. The superstar companies are the exception, not the rule - and they are pulling further ahead every year.
Not everyone reads this data with the same optimism. The report itself, in a footnote, references the Economist's recent warning to "prepare for an AI jobs apocalypse" - a reminder that PwC's analysis of job advertisement data captures what employers are asking for, not necessarily what is happening to the people whose roles are being quietly restructured beneath them. The professionalised jobs are real and growing. So are the democratised ones whose occupants may not yet realise what is happening to them.
The dividing line the PwC barometer describes is already running through insurance. The question now is which side of it individual firms - and individual professionals - are on.
