Rewiring Roles: What HRBPs Need to Know About AI and the Workforce Shift
The Hot-Potato Problem: Who Owns the AI Skills Gap?
TL;DR: AI isn’t replacing employees, but it’s reshaping the roles we hire for, develop, and promote. HR Business Partners (HRBPs) need to shift from talent firefighting to proactive design thinking to keep up. The AI skill gap is real, but the bigger issue is figuring out who owns that gap and how to close it. Spoiler: it’s on HR to take the lead, and it starts with reimagining jobs from the ground up.
Let’s start with a scene that keeps playing out in not-so-quiet executive meetings:
“We all agree there’s an AI skills gap – but where does it live? L&D? The business? Resource management? HR?”
Heads nod solemnly around the table. Yes, yes, huge gap, very important. But when it comes time to act, the accountability gets passed around like a hot potato. Everyone acknowledges the gap; no one owns it. This confusion isn’t about awareness – it’s about ownership. In many organizations, AI and skills strategy have no single owner. One industry expert noted that at many large companies, different HR or business teams tackle bits and pieces of the skills puzzle with “no governance model, no real ownership, and no dedicated team” leading the charge. In other words, lots of well-meaning activity, but no central steering wheel.
Meanwhile, the clock is ticking. The “AI skills gap” we talk about isn’t a static checklist item for HR to eventually get to – it’s a fast-expanding void between the skills our workforce has and the skills our workforce needs next. And nowhere is that void more evident than in the realm of agentic AI skills – things like prompt engineering, tool orchestration, AI reasoning, and critical evaluation of AI outputs. Not only is the talent pool for these skills shallow, but the skill set itself is brand new. You can’t pull a seasoned ChatGPT strategist with 10 years’ experience out of LinkedIn – that job didn’t exist two years ago. Heck, 87% of executives admit they’re struggling to find talent with AI skills at all. So if we keep punting ownership of the problem, we’ll keep struggling to find people who can actually leverage these emerging tools.
AI at the “Speed of Panic”: The New Reality of AI-Powered Work
Let’s talk about why this is urgent. In just the past few weeks, we’ve seen an AI avalanche of updates and product launches that make one thing clear: these aren’t fringe experiments anymore – they’re enterprise-ready and moving at the speed of panic. Consider just a sampling of recent news:
OpenAI rolled out more autonomous AI agent capabilities for ChatGPT. It’s no longer just autocompleting your sentences – now it wants to autonomously handle tasks like running your calendar or setting up team meetings.
Salesforce has been rapid-fire releasing Agentforce updates, seemingly faster than most companies update their OKRs. (In fact, Salesforce’s Agentforce platform for AI agents now pushes out new features as often as monthly.) The result? Thousands of companies have jumped on board. AI agent usage is up 233% in just six months, according to Salesforce, and 8,000 customers have signed up to deploy AI agents in that time. In one case, a company’s AI agents autonomously resolved 70% of customer chats during their busy season – these tools are delivering real value, fast.
Google is integrating AI “agents” everywhere, too. They launched features like Agentspace and Agent2Agent for Google Workspace – basically letting Gmail, Docs, Calendar, etc., talk to each other and take actions on your behalf (hopefully without unionizing next).
Microsoft has woven its Copilot AI into the fabric of Office and Windows, effectively making it the default “enterprise agent” for millions of users. Whether those employees know how to use it effectively yet is another question entirely.
Workflow automation platforms like n8n are saying, “Hey non-techies, want to build your own AI intern? Here’s the kit.” In other words, even no-code tools are making it drag-and-drop simple to create little AI helpers to automate tasks.
Major enterprise software providers – Workday, ServiceNow, Atlassian, and others are all baking AI copilots and agents into their platforms. Startups are popping up left and right (Cognosys, Orby, MultiOn, Synthflow, to name a few) to make multi-agent workflows feel like second nature.
In short, AI isn’t coming soon; it’s here now, spreading through every function at work. These systems are increasingly powerful and increasingly present in the day-to-day flow of work. It’s telling that 92% of companies plan to increase their AI investments over the next three years, and almost all big organizations are experimenting with generative AI in some form. We’ve entered an era where AI can do more than just generate a pretty paragraph – it can take action. This new class of “agentic AI” means AI systems that don’t just answer questions, they execute tasks. And that has a cascading impact on the skills people need to work alongside these systems.
Here’s the implication for HR: Skills that might have seemed niche a year ago (like crafting a good prompt for an AI, or knowing how to “program” an AI agent through natural language) are rapidly becoming foundational to how work gets done. These are not “nice to have” extras in a job description – they’re becoming core competencies for roles across the board.
Not a Sourcing Problem – a Role Evolution Problem
So you dig into the talent pipeline for these new skills. You come up empty-handed – not much out there externally, and not much internally either. That’s when it hits you: this isn’t a mere recruiting problem, it’s a role evolution problem. AI isn’t just changing the tools we use; it’s fundamentally altering how work gets done and even what “qualified” means for a given role.
Think about what’s shifting under our feet:
Many repetitive tasks are becoming obsolete. Data collection, basic data crunching, routine report generation, rote compliance checks – if it’s tedious and follows a pattern, there’s a good chance AI can automate a lot of it. By 2030, about 40% of core skills in jobs are expected to change as automation and AI handle more tasks. Administrative roles are already seeing this effect – no surprise that roles like data entry clerks, executive assistants, and bank tellers are among the fastest declining jobs as AI and automation advance.
New tasks (and even entirely new jobs) are emerging. We’re seeing roles like “Prompt Engineer” or “AI Ethicist” spring up, and existing jobs are getting AI-related twists (e.g., Marketing Managers now need to understand how to get the best out of an AI copywriter). A year ago, people scoffed at the idea that “prompt engineer” could be a six-figure job; now, some companies have offered $300k+ salaries for those skills. That’s how valuable AI proficiency has become overnight.
Soft skills are center stage. When you strip away the routine busywork from a job, what’s left are the distinctly human aspects, such as judgment, creativity, communication, mentoring, and empathy. In fact, employers say the top skills they need now are analytical thinking as well as resilience, flexibility, and leadership – the kinds of human skills that machines can’t easily replicate. As AI handles more of the grunt work, an employee’s ability to exercise sound judgment and communicate effectively about AI-generated outputs becomes a critical differentiator. (AI might draft the report, but you need human judgment to decide if the report actually makes sense.)
Job descriptions and org charts aren’t keeping up. Many organizations are still writing job reqs and backfilling roles like it’s 2018, ignoring that the role’s daily tasks may have radically changed in the past year due to AI. The market hasn’t caught up – you’ll see job listings asking for years of experience in technologies that barely existed 18 months ago, and on the flip side, roles that should require AI skills often don’t even mention them. There’s a lag in our talent frameworks. Most orgs refresh their job architectures about as often as they update the office emergency exit maps (read: rarely). But if roles are evolving faster than they’re being redefined, we’re flying blind.
Let’s make this concrete. Consider the humble Data Analyst:
Five years ago, a Data Analyst’s job might have included gathering data from various sources, cleaning up spreadsheets, producing routine weekly dashboards, and writing up summary reports.
Today, an AI-augmented Data Analyst can have an AI agent gather and clean the data, auto-generate initial charts or dashboards, and even draft insights. The Data Analyst’s new tasks are higher-level: validating the AI’s outputs for accuracy, designing the right prompts or queries to get meaningful analysis, coaching business stakeholders on interpreting the dashboards, and ensuring the AI isn’t hallucinating false trends. In other words, less time on grunt work, more time on quality control, interpretation, and strategic questioning.
It’s the same job title, but the role anatomy has been rewired. Old tasks offloaded, new tasks added, and the skill profile shifted from technical data wrangling more toward critical thinking and consultative skills. This story is unfolding across multiple roles within the organization.
Yet, too many companies are still trying to backfill the old version of the role. They haven’t updated the job definition to reflect these new realities. When “what good looks like” in a role has changed, but your job descriptions and talent criteria haven’t, you end up either hiring the wrong people or underutilizing the right people.
What’s the bottom line? AI is not “replacing jobs” outright; it’s rewriting them. Research backs this up: a recent study estimated that 80% of workers will have at least 10% of their tasks affected by AI, and 19% of workers could see half or more of their tasks automated in the near future. But those tasks don’t equal whole jobs. In most cases, AI will supplement jobs, not completely supplant them. The work gets redistributed, not entirely eliminated. Some roles will indeed disappear, but many more will morph, and new ones will appear. The net outlook isn’t one of mass unemployment – it’s one of mass redefinition. The World Economic Forum predicts a net 78 million new jobs will be created by 2030, even after AI-driven losses, as jobs are transformed rather than extinguished.
From Knowledge Worker to Superworker
If AI is rewriting jobs, who wins in this new world? Enter the “Superworker” concept. Coined by HR industry expert Josh Bersin, a Superworker is basically a knowledge worker on AI steroids – an individual who leverages AI to dramatically enhance their productivity, performance, and creativity. These are the employees who don’t fear AI will replace them; they figure out how to make AI work for them. In Bersin’s words, “The rise of the Superworker depicts AI as a co-worker, enhancing people’s capabilities rather than rendering them obsolete”. In other words, AI isn’t the rival; it’s the sidekick.
Think of it this way: take an average employee, automate 30% of their drudge work, give them on-demand insights and suggestions via an AI tool, and what do you get? Someone who can devote more time to high-impact activities, the strategic, creative, human parts of the job. In the story above, our Data Analyst Jill, who started using AI to automate the busywork, became a strategic advisor, identifying trends and advising leadership, instead of just cranking out reports. Her managers suddenly wish they “had three more Jills” on the team. Jill became a Superworker, and her company reaped the benefits.
This isn’t just feel-good rhetoric; companies that embrace Superworkers are already pulling ahead. Bersin talks about “Superworker companies” – organizations that redesign work and invest in AI augmentation at scale. Early evidence suggests these companies (sometimes called AI Pacesetters) far outperform their peers. Why? Integrating AI into daily tasks enables employees to focus on higher-value work, thereby boosting efficiency and innovation. One Salesforce client, for example, cut average case handling time by 15% by using AI agents, and another saw a 22% increase in customer retention. That’s the kind of jump you get when people and AI are working in tandem, each doing what they do best.
Crucially, this augmented future depends on learning. The only difference between an average knowledge worker and a Superworker is the ability to effectively use AI tools. And that is a learned skill. Yes, a skill – meaning it can be trained and developed. The gap right now is that most employees haven’t been trained, and most companies haven’t offered training. The result: a whole lot of AI potential sitting untapped. In one survey, 79% of executives admitted that if they personally don’t learn how to use AI, they won’t be prepared for the future. And yet, internal AI education is in woefully short supply. The best way to ensure AI doesn’t make you or your team obsolete is education and training, but “both are in very short supply internally at most companies”. This is a fixable problem if HR and L&D step up to provide that training, you turn apprehensive employees into confident Superworkers. It’s that simple (and that challenging).
The mindset shift we need is to view AI as an empowerment tool, not a threat. When we give employees the training, permission, and encouragement to use AI, we see the fear turn into excitement. Three out of four employees say they’d be excited to use AI at work if their company were more transparent about how AI would improve their workflow. People are largely on board to become superworkers – they just need the support and clarity from leadership.
So, HRBPs: rather than anxiously monitoring if AI will cut jobs, focus on how it can augment jobs. The goal is to turn every knowledge worker into a Superworker, one AI-assisted task at a time. And doing that requires a deliberate game plan. Let’s talk about what that entails.
HRBP Game Plan: 5 Moves to Redesign Work in the AI Era
Most HRBPs are used to putting out fires – the urgent talent vacancy here, the retention crisis there. But now is the time to play offense, not defense. Proactive design thinking is the name of the game: proactively redesign roles, career paths, and skill development before you’re forced to. Here are five moves HRBPs should start making now to future-proof their workforce:
1. Zoom Out and Re-Map Roles
Stop viewing job postings as static checklists of duties; start looking at them as living, breathing snapshots of how work gets done. Sit down with business leaders and map out, task by task, how key roles have changed (or could change) due to AI. Ask: What tasks in this role are now AI-assisted or AI-performed? What new tasks have emerged? You might discover, for instance, that your marketing team’s Social Media Manager is no longer spending hours drafting posts (AI can do 80% of that), but is now spending more time curating and fact-checking AI content, engaging with audiences, or running experiments. That means the skills and priorities in that role need to be updated.
By zooming out and mapping the before and after of a role, you can update job descriptions to reflect reality. This also helps identify redundancies (tasks AI can handle entirely) and new gaps (areas where human oversight is now crucial). It’s a bit like drawing a map of a city that suddenly got new highways built – you need to redraw the streets and update the traffic rules. If you skip this step, you end up trying to hire from a nonexistent talent pool (looking for people great at tasks that are now automated) and overlooking internal talent who could take on the new tasks with some training.
Pro tip: Don’t do this in a vacuum. Involve the people actually doing the jobs they often know exactly which parts of their work have been quietly taken over by AI and which new responsibilities they’ve picked up. Their pain points will tell you where AI is causing friction and where it’s freeing up time.
2. Use a Skills Lens (Not Just Titles)
We’ve all been guilty of a little tunnel vision on job titles and career ladders. The AI era demands a different lens: skills. Specifically, focus on emerging skills, expanding skills, and expiring skills in your organization. Prompt design, AI tool selection, data literacy, and critical thinking – these are emerging as must-haves. Meanwhile, some traditional skills might be becoming less critical (e.g., manual data entry, rote memo writing). And some human skills (like coaching, cross-functional collaboration, and creative problem-solving) are expanding in importance as routine work automates.
Ask yourself and your business leaders: What capabilities do we need more of in the next 2–3 years? What skills are now mission-critical that barely mattered five years ago? Chances are, competencies around AI fluency will top the list. In the World Economic Forum’s 2025 report, nearly 45% of employers said that AI and big data skills are now considered “core” to their business. And that was a global, cross-industry stat – for many knowledge-worker-heavy companies, it’s arguably even higher.
Using a skills-first approach means you might discover talent in unexpected places. Maybe someone in Audit has developed amazing prompt-engineering skills, building an AI tool for personal use, could that be applied in an Innovation or Analytics team project? In a skills-driven organization, you’re looking beyond someone’s current role or title and seeing an adjacent skill that could be a bridge to a new role. When you start tagging and tracking skills (through assessments, skills inventories, or even AI-based skill inference tools), you create visibility: who knows what, and who could learn what next.
This approach also influences hiring. Instead of replacing John the Analyst with another Analyst with the exact same background, you might hire someone from a completely different field who has the core analytical skills and the AI know-how your team now needs. The old playbook of “5 years of experience in X industry with Y software” might not serve you when that software can now be learned quickly with AI help, or is obsolete. Instead, maybe you need “demonstrated ability to learn and work with AI tools in context.” That’s a different mindset, and it opens up your talent funnel to non-traditional candidates who could become your next Superworkers.
3. Partner with L&D to Upskill at Scale
Once you’ve identified those must-have skills, learning and development becomes your best friend. If 49% of skills in the workforce may become irrelevant in just two years (as one survey of executives suggested), the only way to keep up is to learn quickly and continuously. Unfortunately, most companies’ L&D programs haven’t kept pace with the AI curve. (Raise your hand if your corporate LMS still touts courses from 2017 on “Digital Transformation” with a stock photo of a robot, yeah, time for an update.)
HRBPs should champion a serious AI upskilling agenda. This isn’t just a one-off workshop on “ChatGPT 101”. We need ongoing learning journeys that help employees build AI literacy, practice with real tools, and apply new skills on the job. Whether it’s formal courses, hackathons, online nano-degrees, or lunch-and-learns, make AI training a continuous offering. The goal is a workforce that’s AI-confident and AI-competent.
A few pointers on this partnership: First, co-create the curriculum with input from business units. What AI use-cases are emerging in marketing vs. finance vs. HR? Tailor learning to be relevant to each context (people tune out when training is too generic). Second, blend technical skills with human skills. For example, teach prompt engineering alongside critical thinking (knowing how to question an AI’s output). Teach AI-driven data analysis alongside storytelling with data. The human-AI interplay is key.
Most importantly, signal from the top that AI skills are a priority. Leaders should be first in line to get trained – it shows everyone that this is serious. Remember, lack of education is one of the biggest gaps right now. Closing it is one of the few variables we can actually control in this chaotic technological race. As one report bluntly put it, “make it a priority to build or buy formal AI education ASAP”. If you don’t have the in-house expertise, bring in a partner or send folks to external bootcamps. The investment will pay off in spades when you’re not scrambling to find $300k “AI whisperers” on the open market because you’ve grown your own internal talent.
4. Encourage Safe AI Experimentation
Here’s a secret: your employees are likely already fiddling with AI tools on their own. (McKinsey found employees were three times more likely than leaders assumed to believe AI will replace part of their work, and many are already using AI regularly in their jobs.) Instead of a culture of quiet, shadow AI experiments, create a culture of open, safe experimentation.
What does that mean? It means giving teams the green light to pilot AI in their workflows without fear of punishment if something goes awry, within sensible guardrails of ethics and data security, of course. Encourage each department or team to designate an “AI Ambassador” who tries new tools and shares insights. Set up cross-functional forums (virtual or IRL) where people can demo a cool AI use case they tried, or even where teams can hold friendly competitions (who can save the most hours using an AI hack this quarter?).
By normalizing experimentation, you achieve a few things. One, you accelerate learning – people learn from each other’s successes and mistakes. Two, you surface the best ideas to scale up. Maybe one sales team figured out an AI hack to write personalized client emails that doubled their outreach – that’s something to roll out to all sales teams. Three, you build “muscle memory” for adapting to new tech. If employees feel they have the agency to try things, they become more adaptable in general. That’s exactly the kind of culture you need when technology is evolving faster than your annual strategy deck.
A practical step here is to use AI in HR processes as a sandbox. For example, try using AI to draft an onboarding schedule for a new hire, or to summarize employee feedback from surveys, or to help write a first draft of a job description (then let a human refine it). By dogfooding these tools in HR, you not only gain efficiency, but you also role-model the behavior for other teams. It sends a message: it’s okay to try AI, and here’s how we’re doing it responsibly.
Of course, ensure there are guidelines, e.g., don’t paste sensitive data into public AI tools, verify AI outputs, etc. “Safe” experimentation is key; we’re not just unleashing chaos. But a bit of controlled chaos can be healthy. Remember, innovation rarely comes from total comfort. A little discomfort and risk-taking, in a safe environment, will prepare your workforce to adapt when bigger shifts come down the line.
5. Push for Refreshed Job Architectures
This one might be the most abstract, but it’s hugely important: modernize your job architecture. By job architecture, we mean the whole system of job families, leveling, career paths, and how roles relate to each other in the organization. Many companies treat their job architecture like a dusty artifact – built for a world of static hierarchies and linear careers, then left on a shelf. In the AI era, that’s not gonna fly.
We need job architectures that are flexible, matrixed, and skills-oriented. Roles are no longer endpoints or fixed slots in an org chart; they’re more like springboards or nodes in a network. You might hire someone into Role X in Team Y today, but if that person has the ability to contribute in Z way next month (thanks to new skills or an AI shift), you should be able to redeploy them without a year of bureaucracy. The question to ask is: Does our job framework enable fluid movement and evolution, or does it put people in silos?
Consider a few markers of a future-ready job architecture:
It enables capability mapping across functions. (E.g., you can find all the “data visualization” specialists across different departments, regardless of their title, because you’ve tagged that skill in your system.)
It highlights adjacent skills and stretch potential. (E.g., an employee profile doesn’t just list what someone has done, but what they could do next – maybe your Finance analyst also has a certification in UX design or a knack for prompt engineering, and that opens up non-linear career moves.)
It has a common skills language underpinning it. (This might mean adopting a skills taxonomy or using AI to infer skills, so that when one part of the business says they need “machine learning knowledge,” you can connect it to people or roles elsewhere that have that, even if they’re in an unrelated-sounding job title.)
The benefit of refreshing job architecture is that it becomes much easier to see the gaps and the opportunities. If AI has reshaped tasks, you’ll likely discover that some job families need a new branch, or some levels need new definitions. For example, maybe you add a new level in the Software Engineering job family that specifically recognizes expertise in AI tool integration. Or you realize that the Marketing and Data job families have an overlap now in the area of “AI prompt writing” – so you adjust your frameworks to account for a role that bridges that gap.
Organizations are already moving in this direction. We see deep hierarchies giving way to flatter, project-based structures, and traditional jobs giving way to more modular roles or gigs. As one skills strategy expert observed, “deep organizational hierarchies and complex job architectures are giving way to flatter, team-based structures, and leaner, more flexible job architectures.” The rise of internal talent marketplaces at many companies is a testament to this – instead of locking people into a narrow job description, companies like Unilever, IBM, and others are using platforms to match employees to projects or gigs based on skills. That’s a skills-based job architecture in action.
HRBPs can be the champions of this change by pushing for a cross-functional task force to revisit job definitions. Work with Comp, with Talent Management, with business unit reps. Yes, it’s a project, and yes, everyone’s busy – but this is foundational work that will make everything else (hiring, development, mobility) smoother. If your company’s roles and leveling still assume a pre-AI division of labor, you’re effectively managing to a world that no longer exists. As the saying goes, structure drives behavior – if we want people to become continuous learners and move fluidly to where they’re needed, we need structures that support that.
One more thing: transparency. Modern job architecture should make skills and pathways visible to employees. In a dynamic environment, you want employees to see, for example, that someone in Audit could indeed transition to an Innovation role if they build on their data skills with some AI experience. Or that a customer support rep could become a “CX analytics lead” because they’ve learned to leverage AI in understanding customer queries. When people see non-traditional moves are possible, they stretch themselves. If we keep them in the dark with rigid org charts, they’ll stay in their lane – and that’s a recipe for stagnation when we need agility.
Zooming Out: HR’s New Mandate in the Age of AI
The big picture in all this is that HR’s role is shifting from managing talent to designing work. In an AI-reshaped world, we can’t just play musical chairs with employees (hire, fire, replace) and call it a day. We have to redesign the chairs. HRBPs, sitting at the intersection of people and strategy, are uniquely positioned to guide this redesign. It’s time to move from being reactive problem-solvers to proactive futurists for our organizations.
This means embracing a bit of a designer’s mindset: be curious, experiment, prototype new role definitions, solicit feedback, iterate. It also means being a translator between the tech and the people. AI will throw up a lot of jargon and fast-moving developments; HR can translate that into what it means for skills, for jobs, for organizational culture. If leadership isn’t sure how to steer, bring data and research: show them that companies who boldly adopt AI (and upskill their people) are outperforming, show them the risk of doing nothing (half of your workforce’s skills could be outdated in two years, remember). In McKinsey’s 2025 study, they found employees are ready for AI and it’s often leadership dragging feet. HR can help close that gap by informing, educating, and sometimes nudging hesitant leaders to take action.
Finally, keep it human. Ironically, the more we infuse AI into work, the more the human touch matters. Culture, empathy, purpose – these become even bigger differentiators when technology levels the playing field on technical tasks. HRBPs should advocate for balancing efficiency with empathy. Automate the drudgery, yes, but use those saved cycles to invest in career conversations, mentorship, and well-being. AI can crunch numbers; it’s on us to double down on the people side of the people function.
The future of work is not a fixed destination; it’s a moving target. But one thing is clear: we can’t just backfill and backpedal our way into the future. We have to deliberately create new roles, new skills, and new organizational models that harness the best of AI and human potential. HRBPs, this is your moment to step up as architects of that future. The companies that thrive will be those that create Superworkers and even Super-teams, not those that treat people as replaceable widgets. And creating superworkers isn’t about hiring unicorns – it’s about enabling your existing people to ride the AI wave with skill and confidence.
So next time that exec meeting rolls around and someone asks, “Who’s owning the AI skill gap?” you know the answer (psst, it’s you, HR). Own it proudly, because the work you do now to rewire roles and reimagine skills will decide whether your organization sinks or soars in the coming years.
Your turn: What’s one role in your org that has already been quietly rewritten by AI – title unchanged, but the job today is nothing like it was before? Please drop me a note or comment. I’m collecting examples for a follow-up piece, and I’d love to hear real-world stories of roles in flux.
(The views expressed here are mine, not my employer’s. Let’s lead the change.)
Sources: