Australia's insurance industry has rarely faced a workforce challenge as acute as the one now staring it down from both ends of the career ladder. The Insurance Council of Australia has warned that nearly 30% of the current workforce will be at retirement age by 2030, and according to a Gallagher Bassett white paper, 92% of Australian insurers say a shrinking talent pool is already impairing their ability to manage claims and grow their businesses — well above the 72% global average.
Meanwhile, from the other end, junior hiring is contracting. The entry point into the profession is narrowing at exactly the moment when the profession most needs a strong pipeline of people coming through.
The dominant explanation in most boardroom conversations is generative AI. Tools that can draft policy documents, process claims data, and summarise complex risk assessments are assumed to be doing the work that junior staff once performed, making the hiring of graduates and trainees economically harder to justify. Rod Fitzgerald, managing principal of LimeBox Consulting, has flagged implementing effective talent acquisition and retention strategies as one of 2025's biggest issues for brokerages, in a context where AI is frequently cited as restructuring the underlying work.
But a major new working paper — published in May 2026 by researchers at the University of Warwick, the London School of Economics, and Oxford's Ellison Institute of Technology — challenges that account directly. Drawing on 243 million new hire records and 407 million job postings across four Anglophone economies between 2017 and 2025, the study finds that when the effects of AI exposure and working-from-home exposure are properly separated, the WFH effect is robust and the AI effect largely disappears. The paper's title — The Broken Ladder — is an apt description of what Australian insurance is experiencing.
Junior hiring decline, 2017–2025
Junior share of new hires has fallen sharply across all four countries
Percentage-point change from 2019 baseline — United States, United Kingdom, Canada, Australia
Dashed lines mark COVID-19 onset (Q1 2020) and ChatGPT release (November 2022). Series are quarterly, seasonally adjusted, reweighted to hold occupation mix constant at the 2019 US distribution. Source: Lambert & Schindler (2026), The Broken Ladder; Revelio Labs.
The numbers from Insurance Business Australia's own reporting make clear this is not a recent phenomenon manufactured by language models. More than 90% of insurance firms surveyed in Australia said a limited talent pool is affecting business growth and claims management, compared to 72% of respondents globally, according to Gallagher Bassett's Carrier Perspective: 2025 Claims Insights report. With about 30% of the industry's professionals expected to retire within the next five years, concerns are growing, and the smallest brokerages are bearing the brunt, unable to compete with larger networks for a dwindling pool of early-career candidates.
Talent attraction and retention have moved up the priority list for Australian insurers, rising from seventh place in 2025 to third in the latest Gallagher Bassett Carrier Perspective: 2026 Claims Insights report, with three in four Australian insurers reporting greater difficulty in recruiting suitably skilled staff — with the greatest shortages in claims management, loss adjusting, and specialised case management. The gap is not simply about headcount. It is about the erosion of the tacit knowledge, the relationship capital, and the professional judgment that insurance work demands and that cannot be acquired from a manual.
The paper, The Broken Ladder: AI, Remote Work, and Early-Career Hiring, by Peter John Lambert and Yannick Schindler, makes a finding that is directly relevant to insurance employers. Both AI exposure and WFH exposure, when measured in isolation across occupations, appear to predict falling junior hiring with roughly equal force. The problem is that both measures rank the same occupations — knowledge-intensive, computer-heavy, analytically demanding white-collar roles — near the top of both indices. Insurance is almost exactly the archetype of a WFH-capable and GenAI-exposed profession.
When both variables are entered into the model simultaneously, the WFH coefficient remains large and statistically significant. The AI coefficient collapses — often to near zero. This is not a marginal finding. A confounder explaining only 1–3% of residual variation would be enough to eliminate the AI effect entirely. Eliminating the WFH effect would require a confounder five times more powerful. The authors also test the result using actual observed WFH adoption — which firms explicitly offered remote or hybrid roles in 2021–22 — and find the same result holds. WFH adoption predicts lower junior hiring in the years that follow.
For insurance, the mechanism is especially intuitive. The paper's model shows that WFH raises the cost and reduces the return of investing in junior talent by disrupting the proximity-based knowledge transfer that those investments depend on. Insurance — particularly broking, underwriting, and complex claims — is a craft profession. It is learned by watching how a senior broker handles a difficult client, how an underwriter weighs the grey areas in a complex risk, how a claims manager negotiates with a loss adjuster. Nick Beswick, head of professional indemnity for Newline Australia, has noted that gathering teams for substantive discussions becomes genuinely difficult when everyone can be in different locations — a friction that bears disproportionately on the newest members of any team.
The industry has recognised the mentorship problem and is responding, but the responses are uneven. Gallagher Bassett fosters an internal mentoring program, with claims operations manager Lisa Odgers describing mentoring as a way to build future executives. Canopius Group has a well-established mentorship scheme, with head of Australia and Pacific Claudio Saita describing mentoring as essential to developing and training staff to create more rounded and high-performing employees. Mentorships are also central to NIBA's talent development efforts, supported by ANZIIF's Careers in Insurance programme, with industry leaders noting that these programmes help employees pursue productive careers and improve inter-staff relationships.
The industry's Rising Stars 2025 profiles illustrate precisely what is at stake. One young broker described going from the most junior team member to running the liability function solo, forced to lean on other parts of the business and work with executives on strategic initiatives — learning by immersion in a way that distributed work arrangements make structurally more difficult to replicate. Another rising professional credits a NIBA mentor programme for shaping her leadership journey.
These are not incidental outcomes; they are the product of the physical proximity and informal observation that hybrid work disrupts.
The rising Stars research also points to changed expectations of junior professionals. Junior insurance roles that were once task-based and time-bounded have expanded: junior staff are now expected to lead cross-functional projects, manage vendors, and take on ESG or digital transformation work much earlier in their careers. These are demands that require judgement developed through exposure — not skills that can be transmitted via a Teams call.
Insurance is not homogeneous on working arrangements, and that heterogeneity is part of the problem. Three days in the office and two at home has emerged as an informal industry standard for many Australian insurers and brokers, with some larger firms mandating full-time office attendance while others, particularly family-focused brokerages, have settled into the 3-2 pattern. Recruiters have warned that firms insisting on five-day office attendance risk struggling to attract talent in a market where hybrid work has become a baseline expectation.
This creates a genuine dilemma that the Lambert-Schindler paper does not resolve but usefully clarifies. The dilemma is not between office and remote work per se — it is between what firms need to do to attract experienced candidates (offer flexibility) and what firms need to do to develop junior talent effectively (provide proximity). Most hybrid policies apply the same framework to a senior underwriter of fifteen years' experience and a graduate in their first six months. The paper's evidence suggests these are fundamentally different management situations, and treating them identically may be one of the mechanisms driving junior hiring down.
Gallagher Bassett's approach to shifting focus toward long-term talent strategy reflects a broader industry movement: investing in training, career pathways, and alternative talent pipelines to address structural workforce challenges. The Insurance Council of Australia has repeatedly highlighted skills shortages as a key issue, particularly in claims handling following major catastrophe events. But investment in pipeline development is undermined if the organisational conditions that make junior development tractable — physical proximity, informal feedback, visible performance — have been removed by the same hybrid arrangements that firms are offering as recruitment incentives.
The Lambert-Schindler paper is explicit that its findings are cause for optimism rather than despair. If WFH, rather than AI substitution, is the primary driver of declining junior hiring, the problem is organisational — and organisational problems can be solved. Several practical steps follow directly from the research and from the Australian insurance industry's own experience.
The first is diagnostic. Audit whether your firm's junior hiring decline tracks most closely with the business units that went hybrid earliest, or with those deploying AI tools most aggressively. For most Australian insurance operations, the answer to that question is likely to be more informative than any technology benchmarking exercise.
The second is structural. Design hybrid policies that explicitly differentiate between experience levels. The expectation of physical presence for early-career staff — say, in the first two years of tenure — is defensible and widely accepted by younger workers who understand the development rationale. Presenting this as an investment in their career, not a restriction of their flexibility, changes the framing from punitive to developmental.
The third is deliberate. The efforts that need ramping up include mentoring programs, university partnerships and school outreach campaigns, including the ICA's Insurance Careers initiative, according to NIBA's 2025 Queensland Insurance Broker of the Year. Mentoring cannot substitute for physical proximity, but structured, deliberate mentoring programmes — with explicit time commitments from senior professionals — can compensate for the incidental learning that hybrid work removes.
Australia's insurance industry is at a critical demographic juncture. The senior professionals whose knowledge and relationships sustain the industry's quality of service are approaching retirement. The junior professionals who would normally be absorbing that knowledge through proximity, observation, and informal apprenticeship are harder to hire and harder to develop under the working arrangements now common across the sector. That is not an AI problem. It is a management problem. And in the insurance industry, management problems have always been solved by the same mechanism that builds great brokers, underwriters, and claims managers: one person, in the same room, showing another person how it is done.
Peter John Lambert is at the University of Warwick and the London School of Economics. Yannick Schindler is at the Ellison Institute of Technology, Oxford. The Broken Ladder: AI, Remote Work, and Early-Career Hiring was circulated in May 2026. Insurance-specific data sourced from Gallagher Bassett Carrier Perspective 2025 and 2026, Insurance Council of Australia, NIBA, and cited Insurance Business Australia reporting.
