When 76% of insurers have deployed AI but half the workforce is looking to leave, the technology strategy and the talent strategy are not the same strategy.
getty
The insurance industry’s AI transformation is real and accelerating. Claims automation, algorithmic fraud detection, underwriting models processing risk at scales no human team could match. Deloitte’s survey of 200 US insurance executives finds that 76% have already implemented generative AI in at least one business function. Yet the gap between implementation and scaled deployment remains the defining execution challenge of the moment. The efficiency gains where scaling has occurred are measurable and the competitive logic is sound.
The industry needs them. P&C insurers are caught between margin pressure, climate-driven loss volatility and cost inflation pushing combined ratios toward breakeven. Life insurers face their own version: legacy system drag, rising claims volatility and persistent pressure on investment returns. In that context, automating what can be automated isn’t optional. It’s survival.
But AI isn’t just automating tasks. It’s dismantling the developmental ladder insurance careers were built on. Junior adjusters learned by handling routine claims. Entry-level underwriters built judgment by working through standard risks. Those experiences, repetitive, low-stakes and high-volume, were how people developed the pattern recognition and professional instinct that made them valuable over time.
By Q4 2025, one in three insurers reported at least one AI agent running in production across underwriting or claims functions, according to ScienceSoft’s Insurance AI Trends analysis. The administrative and junior analytical roles being thinned out by automation have long served as training grounds: the way employees learned the craft of underwriting or claims judgment. Without those early rungs, the industry risks losing the apprenticeship model that has sustained it for generations.
ReSource Pro’s 2025 research projects a shortfall of 400,000 insurance jobs by 2035. That number reflects not just displacement but a structural hollowing of the pipeline that produces experienced insurance talent.
What remains, and what becomes more valuable precisely because AI handles everything else, are capabilities no algorithm replicates well. Pattern recognition across complex, ambiguous situations that don’t resolve into clean data. Emotional intelligence in high-stakes conversations where what the customer isn’t saying matters as much as what they are. Creative judgment in unprecedented loss scenarios where there is no model to validate against.
These are not soft skills. They are the core of human contribution in an AI-augmented workforce. And the industry is not developing them. New research from my employer Gallup highlights the talent challenge the insurance industry faces.
Engagement Isn’t the Problem. It’s the Evidence.
Gallup’s latest insurance workforce data shows employee engagement dropped from 39% in 2021 to 36% in 2025, a three-point decline in four years. Close to half of insurance employees are actively seeking new jobs. Only 29% would recommend their organization as a place to work, down from 39% in 2021.
Insurance Industry Gallup Trends
GALLUP
The standard response is a retention initiative. Better compensation benchmarking. Flexible work arrangements. Manager training. These aren’t wrong. But they treat a structural signal as a cultural problem.
What the engagement data is measuring is harder to fix: employees have stopped believing that what’s happening in their organizations is worth staying for. Not because the work is unpleasant, but because the implicit contract between the organization and the workforce has broken down.
That contract was simple: work hard, build skills, advance. It breaks down when AI is rewriting what skills matter and most organizations haven’t decided what to replace them with. PwC’s insurance workforce research, drawing on its 2025 Global AI Jobs Barometer, finds that skill requirements for roles exposed to AI are evolving 66% faster than those in other fields, and over 40% of entry-level employees believe technological change will impact their jobs to a major extent over the next three years.
These aren’t abstract anxieties. They’re employees watching the developmental pathway disappear and looking for evidence that their organization has a plan.
Gallup’s broader data reinforces the diagnosis. Only 1 in 5 employees have clear development plans. Only 2% of CHROs believe their people have the skills their organizations need. The workforce can see this. The 48% actively seeking other jobs aren’t leaving because the work is bad. They’re leaving because they can see the talent architecture isn’t pointed at the future the organization claims to be building toward.
The customer-facing consequences are already showing. J.D. Power’s 2025 Individual Life Insurance Study finds that only 19% of customers describe their relationship with their agent as trusted, while 58% are classified as disengaged or transactional. When agents do deliver trusted service, satisfaction scores are 253 points or higher than relationships that fall short of that standard. On the P&C side, J.D. Power finds that when claims communication is difficult rather than easy, satisfaction scores drop by 440 points — a gap almost entirely determined by the human handling the interaction.
In an industry where the product is a promise and the relationship is the distribution channel, that gap matters. When the people who build and maintain those relationships are walking out, the engagement crisis stops being an HR problem and becomes a business model problem.
The Misalignment Leadership Needs to Own
Deloitte’s research also shows 90% of insurance executives agree on the urgency of reinventing workforce capabilities for human-machine collaboration. Only 25% have taken tangible action. That gap is easier to explain than it looks.
The problem isn’t that insurance leaders don’t understand the urgency. It’s that the systems they’re relying on to execute the transformation were built for a different set of outcomes. Performance management, career architecture and development investment: each was designed for a world where the skills were stable and the roles were predictable. This is a leadership design problem, not a talent execution problem. The systems produce what they were built to produce and no amount of program investment changes that until the architecture changes.
The consequences are visible and measurable. Gallup’s insurance workforce data shows only 23% of employees strongly feel they are managed in a way that motivates them to perform better. Only 22% strongly agree they received meaningful feedback in the past week. Only 23% strongly agree that exceptional performance is rewarded in their company. These aren’t sentiment scores. They are performance system readouts. They describe an industry where the feedback loops that develop judgment have largely stopped functioning, precisely at the moment when developing judgment is the whole job.
PwC’s research on insurance AI implementation names the outcome directly: when AI concentrates expertise in small, experienced groups while automation assumes routine work, it creates fewer opportunities for others to develop the critical thinking skills that foundational work enables. Left unchecked, this produces what PwC calls a “brittle” workforce, one that struggles to step in when AI errs or conditions change. In an industry where a misplaced phrase in a denial letter can be more damaging than a delayed payment, brittleness isn’t an abstract risk. It’s a liability.
The cost compounds in ways that don’t show up immediately on a combined ratio. The experienced claims leader who understands fraud patterns across lines of business. The underwriter who can read a risk that falls outside the model’s training data. The relationship manager who knows when a policyholder is about to lapse and why. These capabilities take years to develop and they’re walking out at a rate the engagement data makes impossible to ignore.
When only 29% of insurance employees would recommend their organization as a place to work, that’s not a culture problem. It’s a design verdict.
What Structural Alignment Actually Requires
The question for insurance leadership teams isn’t how to accelerate existing talent programs. It’s whether the underlying architecture is capable of producing the workforce an AI-changed industry requires. Each of the three systems that need redesigning is a substantial undertaking, but none requires starting from scratch.
Performance Management
Performance management systems need to shift from rewarding volume to rewarding judgment, not by adding a new competency framework alongside old metrics, but by changing what gets measured, what gets recognized and what determines advancement. Organizations that have done this well typically start with a single function or business unit, build evidence that the redesigned system produces different behavior and expand from there.
Career Pathways
Career pathways need to be rebuilt around the capabilities that become more valuable as AI scales, not the roles automation is making obsolete. In insurance that means designing career progression around human judgment, contextual intelligence and relational capability rather than mastery of workflows AI is absorbing. The claims adjuster who becomes a complex case strategist. The underwriter who becomes a risk architect for situations the model can’t price. The path needs to exist and be visible before the talent decides it doesn’t.
Development Investment
Development investment needs to move beyond tool proficiency, which is table stakes, toward building the pattern recognition, emotional intelligence and creative judgment that no algorithm replaces. That shift requires different learning experiences: cross-functional exposure to complex loss scenarios, ambiguous decision-making under real stakes and coaching relationships with senior practitioners who have navigated the transition.
This work cannot be sequenced after the AI strategy is settled.
The talent architecture and the technology roadmap need to be built in parallel, because capability gaps compound faster than development programs can close them if the design work starts late.
The questions worth putting in front of a leadership team before the next AI investment decision:
Are our talent investments preparing people to work with AI or to do what AI cannot and have we defined that distinction clearly enough to build a development program around it?
Do our performance systems reward the capabilities we claim to be developing or the behaviors we’re trying to move away from and where exactly do those two things diverge?
Do our career pathways lead toward the roles we’re building or the roles we’re automating and do our most capable people know the difference?
These questions don’t produce comfortable answers. They reveal gaps easier to leave implicit than to surface in a leadership team meeting.
The Leverage That Already Exists
Insurance leaders have more leverage in this moment than the execution gap suggests. Every efficiency gain, every claim processed faster and every risk assessed algorithmically, makes the human capabilities that remain more visible, more scarce and easier to build a business case around. The argument for investing seriously in judgment, emotional intelligence and complexity navigation has never been easier to make, because the limits of what automation can replace are becoming impossible to ignore.
The challenge is converting that argument into structural change before the engagement decline becomes a talent exodus that’s expensive to reverse. Rebuilding confidence requires more than better programs. It requires demonstrating that the organization’s talent architecture has changed.
The 90% of executives who recognize the urgency are not wrong about the problem. They’re working inside systems that can’t yet produce the solution. Closing that gap is the leadership question that determines whether what insurance organizations are building succeeds or whether they achieve operational excellence while the human capabilities that give the insurance industry its lasting value quietly walk out the door.
