AI Threatens Career Development Pipeline

As AI automates entry-level insurance tasks, insurance industry leaders must design experiences to develop professional judgment in emerging talent.

Insurance

Back in the '90s, I was a beast at reconciling spreadsheets. The 10 fingers I'm using to type these lines used to set my "10-key" numeric keypad ablaze. But if you're among the younger readers to come across this article, you'll have no idea what I'm talking about. And as a society, I think that's for the best.

A core promise of technological advancement is to eliminate drudgery, like using a 10-key to enter vast amounts of numeric data. In property and casualty insurance, we have spent decades trying to remove mundane work from the lives of underwriters, claims professionals, and technologists. We made great strides with Robotic Process Automation (RPA), leaps with broader Intelligent Automation (IA), and now a phenomenal bound with Artificial Intelligence (AI), which promises to revolutionize the way we work.

But amid the terrific change lurks an industry-scale problem. While our job definitions and related skills are morphing at an absurdly fast rate, how do we ensure that emerging talent has the opportunity to develop the depth of understanding and judgment their predecessors possess, even while AI is tending to disassociate humans from some of the traditional activities of their craft?

That replacement system needs a name and, more importantly, an owner. Architecting Experience is the framework of deliberately designed exposures, decisions, and feedback loops through which people develop professional judgment and decision context.

Cracks in the career ladder

Stanford's Digital Economy Lab analyzed payroll data on millions of U.S. jobs through June 2026 and found no economy-wide displacement from AI. What they did find was new employment among 22- to 25-year-olds lagging in industries heavily affected by AI, by about 19% compared to less-affected industries.[1]

The employment ladder isn't broken, but some rungs at the bottom are disappearing, which makes it a whole lot harder to climb.

Even where junior positions aren't disappearing, the formative skills they help develop are the easiest to automate. In insurance, applied experience and apprenticeship have built the tacit knowledge and judgment needed for success for generations. After ingesting enough submissions, you develop a second sense for when a detail looks off, or when risks that look identical stem from entirely different causes. But when this repetitive and seemingly menial work is delegated to AI, younger staff are deprived of the very task-level experiences they need to develop the business EQ, empathy, and scar tissue that will someday make them senior.

What our young talent are left with are the tasks too difficult for AI, and sometimes beyond the judgment they possess. In their 2026 AI Jobs Barometer, PwC looked at over a billion job advertisements. They found that heavily AI-exposed entry-level roles were seven times more likely to demand senior capabilities—like judgment, leadership, and interpersonal skills—than the jobs least exposed to AI. PwC described the trend as the "seniorization" of junior work.[2]

A short-term view of AI-driven task optimization will only make the problem worse. For senior employees, more demanding tasks feel more meaningful and engaging than rote work. But the type of work that is most interesting to a seasoned professional isn't necessarily the type of work best suited to creating one.

And meaningful work sounds wonderful until every task is a demanding one, and the much-needed Easy Tuesday becomes as much of a relic as my most trusty 10-digit keypad.

Clear the slop from the stream

BCG recently surveyed 11,749 workers across 14 markets about their use of AI at work. Two-thirds of regular AI users reported increased job satisfaction. At the same time, 41% reported increased mental strain. Half said they were spending more time managing and directing AI than doing the underlying work itself.[3]

"Brain fry" refers to the mental fatigue associated with information overload from increased AI oversight.[4] With AI, any junior underwriter can generate a 40-page report with the aplomb of a managing director. But then, the report has to be checked. Preferably by someone with enough know-how to identify and correct mistakes.

Without guardrails, we quickly end up in a system where our young talent can access AI-powered tools that allow them to produce work that looks good, but is actually riddled with problems. These problems are then sent upstream to more senior resources who have to spend their time fixing this "workslop"[5]. Junior professionals learn nothing in the process, and increase the workload of their most knowledgeable, and often most expensive, colleagues.

When we claim AI is changing the way we work, we're giving the tool too much credit, while absolving ourselves of responsibility. The truth is, AI isn't changing the way we work. Our decision-makers are. And if, as a result of these changes and the adoption of AI, the leaders break the career ladder, they are responsible for engineering a new one.

AI-resilient architecture

I'm not in favor of clinging to rote work out of nostalgia. AI very much has a place in our workflows, and I remain bullish on the beneficial role of AI over the long term. But today, as we are all still sorting out the impacts and influences of AI on real people, we must think things through holistically and humanistically from the very beginning. In the face of AI removing traditional and lower-level tasks, the implications of this must be assessed, and solutions for the resultant gaps must be created.

When architecting experience for new talent, what matters is to look at tasks through the dual lens of raw efficiency and professional development. Do the tasks teach the rules of the game, and more importantly, when the rules stop applying? Do they offer a chance to shadow a more seasoned professional to develop situational awareness and understanding of context? Do they help build judgment? Do they create the kinds of relationships every business is built on?

These answers will help inform which tasks to automate, and which need more thought. A "yes" doesn't mean the task cannot be handed off to automation; it means the developmental value has to be preserved, somehow. The oversight role for management is to understand and enforce the right order of operations: organic brain first, neural network second.

As with all complex systems, architecting experience isn't a blueprint. It's a process bound to yield different results for different organizations. Once the initial assessment is complete, ideally under the supervision of an executive with accountability, the hard work of execution, review, and adaptation begins. The good news is: there's plenty of responsibility to go around.

Executive management must identify the experiences that create expertise, design a plan for succession, and own how the business will look ten years down the road. Elsewhere, HR and development teams have a responsibility to engineer the learning experiences and coaching or apprenticeship infrastructure, while managers can iterate, suggest improvements, and raise the red flag when skills aren't being developed.

Tech and risk teams also have a clear role to observe AI usage, monitor costs and defect rates, and flag when automated work misses the mark. Senior professionals must keep in mind that juniors need engagement, context, and good professional models to follow, while the juniors themselves must remember to think first and ask AI second.

Insurance is still early in the AI adoption process, and fortunate to have a deep bench of experienced professionals around the world, alongside ambitious newcomers. But every day, AI offers new deployment opportunities for efficiency and optimization. The greatest organizations of tomorrow will not only be able to harness those opportunities, but will also succeed in architecting experience for the next generation of world-class talent.

References

1. Erik Brynjolfsson, Bharat Chandar, and Ruyu Chen. Stanford Digital Economy Lab. "Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of Artificial Intelligence." Revised Aug. 12, 2026. Source

2. PwC. "2026 Global AI Jobs Barometer." June 15, 2026. Source

3. Boston Consulting Group. "AI at Work: Strategy Matters More Than Tools." June 3, 2026. Source

4. Boston Consulting Group and UC Riverside research, reported in Harvard Business Review. "When Using AI Leads to 'Brain Fry.'" March 2026. Source

5. BetterUp Labs and Stanford Social Media Lab. "Workslop: The Hidden Cost of AI-Generated Busywork." Sept. 15, 2025. Source

Read More