The job description is becoming an unreliable contract between an organization and its workforce. Work is changing task by task, while HR systems still update roles in annual cycles.
Sara Braun
Editorial Strategist, HR
Workday
A job description used to be a promise. It told the organization what a role would deliver, told the employee what would be expected of them, and told HR how to hire, pay, and develop the person doing the work. It was the closest thing HR had to a contract between the enterprise and its workforce.
AI is breaking that contract. Jobs have the potential to be fully rewritten, yet the industry hasn’t caught up.
Work is now changing task by task, not role by role. A recruiter's screening work gets absorbed into an AI agent. A financial analyst's variance commentary gets drafted by a model. A customer service rep spends less time answering questions and more time supervising the system that answers them. None of these employees received a new job title over the last quarter. But the actual content of their work has shifted—sometimes dramatically—since the last time anyone looked at their job description.
Meanwhile, most HR systems still treat roles as stable objects that get reviewed on an annual cadence. Job architecture updates lag reality by 12 to 18 months. Competency models get refreshed every few years. Most career frameworks organize jobs into families under the assumption that the day-to-day responsibilities remain static over time.
This gap—between how fast the work is changing and how slowly the role is being updated—has become one of the most consequential blind spots in the enterprise.
Report
Even before generative AI, work had been quietly spilling out from the tidy boxes we drew around it. Employees have been absorbing coordination, translation, and integration work that never appeared on the job posting.
Recent Workday research highlights this reality:
Read those numbers together and a different picture of work emerges. The formal role might say "financial analyst" or "operations manager” or “communications principal.” The actual role is closer to "human middleware"—a person spending a meaningful share of their week bridging the gaps between systems, teams, and data that were never designed to interact.
Infusing AI directly into this environment without restructuring roles only compounds the issue. Gallup research paints a stark picture: only 46% of U.S. employees strongly agree they know what is expected of them at work—down 15 points from a high of 61% in 2015.
"When you overlay AI onto old systems, it just mirrors those old systems rather than changing them," says Vibhas Ratanjee, global leader for leadership development at Gallup. "One big artifact of systems that haven't changed is the job description. In most cases, it remains completely static—people haven't looked at it in years."
The cadence of an annual job review made sense when work evolved slowly. It fails when a single model update can automate a workflow that used to take a team half a day. Consequently, businesses are struggling to adapt: nearly 9 in 10 organizations report that fewer than half of their roles have been updated to include AI-related skills.
Nearly 9 in 10 organizations report that fewer than half of their roles have been updated to include AI-related skills.
HR leaders are already feeling the mismatch in three major places:
Hiring against yesterday's role: Many recruiters recycle postings written before the current wave of AI tools existed. They end up screening candidates for tasks that are being automated in real-time, missing the skills that are quietly becoming essential: critical judgment, system oversight, prompt engineering, and output verification.
Paying for delegated (or absorbed) scope: Compensation bands are typically anchored to defined role scope. When AI absorbs lower-level tasks, employees often take on higher-impact work in return—yet pay structures may lag by a year or more, leaving employees uncompensated for the increased value they create.
Developing stale capabilities: Learning plans assume a clear definition of "good performance." When that target moves faster than the framework, employees notice. Among struggling AI users, 54% say the skills required to do their jobs have not been updated, pointing to a disconnect between formal career paths and real-world demands.
Addressing this shift requires reimagining job descriptions: moving away from rigid task lists toward defining the core business problems a role exists to solve. With AI, the "how" of a job becomes fluid, while the "what" remains the constant anchor.
Rather than trying to hit a moving target, Ratanjee suggests adopting a “Future Success Profile:” a forward-looking, regularly revisited picture of what success in a role requires today and where it is heading. Key points to consider of this new job model include: what work has changed, what work should be automated, and what must remain human-led.
A major hallmark of the traditional job description is what capabilities an employee needs to succeed in a role. That is not going anywhere, Ratanjee argues. But, capabilities need to be bolstered by a philosophy of personal growth and achieving outcomes. “The entire employee experience needs to be connected in a way,” he says. “And it needs to connect with business.”
With AI, the "how" of a job becomes fluid, while the "what" remains the constant anchor.
"A lot of leaders are rethinking how to create clarity—not just telling people what they're expected to do, but showing them how they can grow," Ratanjee notes. "The job description or success profile needs to be evaluated on a quarterly basis to keep up with the pace of technology."
To move toward a living model of job evolution, HR teams can focus on three tactical shifts:
By making intentional changes to adjust to today’s pace of work, the job description becomes a high-level summary of a living thing, not the thing itself.
To adjust to today’s pace of work, the job description becomes a high-level summary of a living thing, not the thing itself.
HR is the only function positioned to oversee the complete picture: the work, the worker, the skills, the compensation, and the ethics. But realizing that potential requires letting go of the traditional annual cycle and the belief that a single PDF can capture modern work.
The job description as a static contract is over. AI is ushering in a much more fluid future, where human potential can evolve right alongside technology. What replaces the job description is a dynamic ecosystem of evolving tasks, skills, and human judgement.
“HR needs to take these artifacts of the past and reimagine them using AI,” says Ratanjee. “I think that’s where we need to go, and that’s what leaders expect.”
Report
A gap between effort and impact is the real ROI problem. Discover the solution.