AI is changing jobs faster than job descriptions can keep up

AI is often framed in terms of what it can do for organisations: boosting productivity, accelerating performance, and increasing revenue. 

But the biggest change is inside the work people do. When using AI, employees spend far less time on execution tasks, like drafting and data entry. They spend more time checking the quality of AI’s output, refining first drafts, thinking critically, and making decisions.

This has a significant impact on what their work actually is compared to what’s in their job description.

Finance teams pressure-test models they didn’t build, rather than building forecasts from scratch. Customer service roles manage the tone and quality of automated responses, rather than type out hundreds of emails by hand. 

These changes are happening in every role, but they aren’t showing up evenly.

A 2025 Stanford study, for example, found that employment is declining among younger workers as AI takes on many of the tasks that make up entry-level jobs. 2025 research by McKinsey found that a third of workers will see more than 10% of their skills change significantly by 2030.

Most of the work HR does across the employee lifecycle is built on assumptions about what roles are. But when that becomes an unknown, HR is making decisions against a document that no longer matches the job.

In a time of AI-driven change, job descriptions as a practice still hold value. They shape candidate expectations, offer employees role clarity, and give HR a compliance guardrail. The problem is that we’re still treating them as a stable record of what roles look like in an environment where fluidity is the new defining factor. 

“We've spent 20 years hiring people for what people have done. AI just made that the least reliable predictor of what they'll do next. We’ve optimised hiring for certainty in a world that no longer supports it.  Job descriptions are already out of date by the time they’re approved. Most of hiring today is built around proxies: years of experience, job titles, company backgrounds. We’ve been rewarding familiarity over adaptability.“

Dominic Joyce

Founder and Head of Talent Acquisition, Maverick Otter

When all bets are off for the capabilities your organisation needs right now, let alone in six months, job descriptions quickly become a task and skills repository. They evaluate candidates based on tools and inputs like “experience with Salesforce”, rather than the cross-functional, transferable skills that show they can adapt when tools and roles change.

This doesn’t just impact who applies for the role — it has a long-term effect on the people already in your organisation, too. 

Because when job expectations shift, they’re no longer a credible frame of reference for other processes across your organisation, such as performance management, development, internal mobility, or progression. Expectations that were once solid slowly drift over time, but get enforced later on.

Data from Personio’s 2025 Workforce Pulse research has found that employees uneasy about AI-driven workforce changes are 29% more likely to look for a new job. 

The bottom line for organisations is that when people aren’t clear about what their role is or how they can grow within it, they leave. 

When roles and skills keep changing, HR teams need a framework for building job descriptions that separates what aspects of the role stay the same versus what changes.

“A lot of organisations are starting with the work and tasks — looking from a lens of what’s stable, what’s more dynamic, and categorising those skills into blocks you can manage. That’s context-driven — mapping your context out and the work that you do. “

Teresa Rose

Director, ConsultHer

In practical terms, HR leaders need to move away from ‘what activities does this person do?’ to ‘what outcomes is this role accountable for?’

“Lock the outcomes, flex the execution,” said Dominic Joyce, founder and Head of Talent Acquisition at Maverick Otter, speaking at Personio's Making AI Work event. “What stays relatively stable would be what success looks like, the level of impact expected, and the responsibilities. What should be fluid would be how the work gets done, what tools are used, and what specific skills are needed.”

Many existing job description frameworks age badly because they try to pin down everything about a role in one document, and don’t give HR teams the flexibility to rewrite the parts that change more quickly than others.

We’ve put together this simple two-layer model to help HR teams make this easier.

The role ‘foundation’ outlines the evergreen things that don’t change about a role, regardless of the tools you’re using, the market you’re hiring in, or the way the work gets done.

Meanwhile, the role ‘flex’ describes the moving parts that orbit that role — the tools, tasks, immediate business priorities, and emerging skills. These are the modular parts of your framework that swap in and out as needs, skills, and tech change.

This framework gives teams a set of principles to help them update the part of the role that changes frequently without rewriting everything else. 

Job description frameworks aren’t just a tool for selling the role and hiring new talent. They’re part of your talent infrastructure — providing the scaffolding for how people are evaluated, promoted, and move around your organisation. 

When AI is reshaping roles faster than you can edit the job description, this scaffolding becomes newly urgent.

“The key question isn't 'is the framework correct?', it's whether it's still helping us make better decisions. Don’t let it sit in isolation. If it's not connected to hiring processes, performance reviews, and internal mobility, it becomes stuck by default.“

Dominic Joyce

Founder and Head of Talent Acquisition, Maverick Otter

Here are the four actions that matter most right now.

Your company has used the same job descriptions for years. Pitching a new framework that changes the way you hire is a hard sell. This is why you need to position it in terms of the risks it mitigates for the entire business.

“Most hiring today is optimised to avoid mistakes, not to find adaptable people. Instead of saying 'we want to move to a skills-based hiring approach', say 'we’re currently hiring from outdated versions of the roles, and that’s creating three risks for us. We’re filtering out adaptable talent. We’re slowing down hiring because we’re chasing perfect profiles that don’t exist. And then we’re increasing mis-hires because we’re hiring for stability in unstable roles'. It’s a way to avoid making the same hiring mistakes at scale.“

Dominic Joyce

Founder and Head of Talent Acquisition, Maverick Otter

When role expectations are changing faster than the documentation can support, an effective career framework provides some stability around what a role is accountable for, what good looks like at each level, and how progression works. But to maintain agility in the current environment, HR leaders need to design for competencies, not tasks. 

This focuses on a value-first role definition without tying roles to a fixed list of tasks that’s likely to change. In the table below are some examples of how that shift is playing out.

A competency-led framework means everyone is on the same page for performance conversations. It also means that in entry-level roles, where tasks are changing quickly, expectations don’t drift or need regular refreshing. At Personio, for example, we’ve added a dedicated AI pillar to our career framework that applies company-wide — treating AI capability as a core competency, not a technical specialism.

This is because a career framework isn’t just for performance conversations — it provides a baseline for the entire employee experience, offering employees a role purpose, scope, and a path for growth even as things keep changing. 

“A career framework really helps keep expectations clear and visible while giving managers a shared language for consistent performance conversations. At Personio, it’s embedded at various points in the employee lifecycle to ensure it’s used consistently — not just when managers are building a job description, but through to performance and how the organisation is shaped overall.“

Suzie Rogers

VP, People Business Partners, Personio

In the current landscape for talent, credentials and past experience have become less reliable measures of how someone will perform. The skills a role needed six months ago might not be the skills it needs now — or six months into the future. 

You can’t hire someone with experience in tools and approaches that don’t exist yet. But you can hire for how someone learns, adapts, and applies judgement in new conditions.

Dominic Joyce said this comes down to four key candidate attributes:

• Learning velocity: How quickly someone can get up to speed on something new

• Transferable thinking: Can they apply what they know to other problems?

• Problem framing: Can they find what needs solving?

• AI collaboration: Can they use AI to enhance how they work, not to avoid it?

“The question shouldn't be who fits the role, it's who adapts when that role inevitably shifts. That is far more predictive for a changing environment than five years doing X, or at company Y. This works because it immediately changes who gets through screening, what you’re evaluating, and how consistent your decisions are — without slowing down anything at all. “

Dominic Joyce

Founder and Head of Talent Acquisition, Maverick Otter

In interviews, HR teams and hiring managers need to move beyond “what tools have you used?” and use questions that reveal learning pace, decision-making, and how someone handles uncertainty.

Instead of: “Tell me about a time when you demonstrated [skill].”

Ask: “Tell me about something you learned faster than expected.”

Add a rubric that assesses: 

• Learning method (how they approached it)

• Feedback loop (how they tested progress)

• Transfer (how they apply the learning elsewhere)

• Ownership (how they handle gaps)

As roles evolve, some skills will naturally rise to the top, while others will become obsolete — much like how we’ve all stopped putting spreadsheets on our CVs. But before you put out an advert for senior engineers with knowledge of some obscure programming language, check your own internal capability first. 

Because sometimes, skills can get trapped in job titles. And this is one of the biggest errors HR can make in the AI era.

“A lot of people in your business right now probably want to move across or up. But you've not created the skills taxonomy to understand who sits where, who's done what. That's why people leave. If you've defined skills properly, you can start to see where those skills exist across the business — internally, where they're needed, but also how to move people between teams.“

Dominic Joyce

Founder and Head of Talent Acquisition, Maverick Otter

To make mobility work, you need:

• Clear data on what roles exist and what they require

• Shared definitions of skills and competencies

• A culture where internal moves are treated as progression, not a sideways step

Reduce role drift quickly without launching a full framework overhaul, by starting small: pick one role family and run a role expectations refresh.

1. Choose one role family where AI is already affecting workflows.

2. Write a one-page role spine: purpose, scope, decision rights, core competencies.

3. Add a short ‘flex’ section: tools/workflows changing now, and the emerging skills that matter.

4. Publish ‘what good looks like now’ in manager-ready language.

5. Set a quarterly check-in (30 minutes) to review what changed and what stayed stable.

6. Add one hiring tweak: one learning-agility interview question + rubric.

7. Make the update visible to employees, not only stored in HR documentation.

This is enough to improve clarity and fairness quickly, and it builds a habit you can scale.

Job descriptions are a symptom of how the world of work is changing. Writing better ones isn’t the goal — but treating them as a mechanism is. Because when done well, job descriptions optimised to match the current pace of change have broader implications for all of your processes across the employee lifecycle. 

Across hiring, performance, development, and internal mobility, they provide the structure that enhances decision-making, offers clarity to employees, and means your people know what’s changing, why, and how they’re a part of it. 

Watch the skills-based hiring and workforce adaptability session from Personio's Making AI Work event, on-demand.

AI is changing the shape of work inside roles. This means employees spend less time on repetitive tasks, such as pulling together research or inputting data — and more time on exercising judgement, quality control, and decision support.

Job descriptions capture a snapshot of a role and responsibilities at the point they’re drafted, and often list tools and tasks that change quickly. When AI alters the tasks the role contains, the gap between the document and the real job can open up in months. 

Start with one role family that is already changing. Define a stable role foundation (purpose, scope, decision rights, core competencies), then add a flex component (tools, workflows, emerging skills) and set a quarterly update cadence.

A career framework defines what a role is accountable for and what good looks like at each level, using competencies rather than a fixed task list. That provides stability even when workflows change.

Anchor performance to outcomes and competencies, then communicate flexible changes early and explicitly. Avoid letting expectations drift informally and only surface at review time.

Skills-first hiring prioritises capability and learning agility over credentials. It helps when you need to fill gaps quickly and when tools and workflows are changing faster than experience-based signals can keep up.