A first week without a job description
Imagine a new employee joining a product team at a fast-growing software company. There is no job description in their onboarding pack.
Instead, there is a dashboard.
It shows the capabilities the organisation needs this quarter: market analysis, feature experiments, customer insight, product documentation and competitive intelligence. AI agents already do much of the underlying work. The team's role is to steer the system, interpret the results and intervene where judgement matters.
By Wednesday, the new hire is working across three initiatives: guiding a research agent through customer interviews, reviewing automated market summaries and shaping the messaging for a new launch.
No single line in a job description could have predicted that work. That is precisely the point.
AI is not simply automating tasks. It is dissolving the structures that used to define them.
The end of the job description
For more than a century, the job description has been the architectural blueprint of the modern organisation.
It emerged from early industrial management to bring order to growing workforces. It defined tasks, responsibilities, reporting lines and required skills, and it became the foundation for hiring, performance reviews, pay structures and career progression. In a world where work changed slowly, the model worked well.
AI challenges the assumptions behind it.
AI rarely eliminates whole jobs in one moment. Instead it absorbs pieces of work across many roles: analysing data, drafting reports, answering questions, reviewing contracts, writing code and organising knowledge. As those capabilities spread, the boundaries between roles blur.
The pace is significant. The World Economic Forum's Future of Jobs Report 2025, based on a survey of more than 1,000 employers, found that they expect 39% of workers' core skills to change by 2030. Tasks move, responsibilities shift and workflows evolve as new capabilities arrive. Static role definitions struggle to keep up.
In fast-moving teams, a job description can be out of date soon after it is written. The document that once defined a role becomes a historical artefact.
Work without fixed boundaries
The deeper shift is in how work is organised.
In AI-enabled processes, work flows through networks of human and machine capabilities. An AI system gathers data, generates insights and drafts recommendations. A person evaluates the result, adjusts the direction and approves the final decision. Tasks move between systems and people depending on context and complexity.
In agentic workflows, where AI agents coordinate sequences of tasks, the design typically pauses for human judgement at key decision points. The human role is not to carry out every step, but to steer the process.
Work becomes more fluid. The central question shifts from who performs a fixed list of duties to who makes sure the intended outcome is achieved.
Responsibility replaces task ownership, and organisational design has to adapt.
From roles to outcomes
For business leaders, the implications are structural.
Most organisations are still designed around roles: marketing manager, financial analyst, customer support specialist. Each carries a defined scope. When tasks start moving dynamically between people and AI systems, that structure becomes brittle.
Some organisations are rethinking the basic unit of design and organising around outcomes rather than roles. Teams own end-to-end goals, such as improving customer onboarding, accelerating product development or making the supply chain more efficient. AI systems do much of the execution, while people oversee the system and step in when complexity or ambiguity arises.
McKinsey's 2025 work on the “agentic organisation” describes where this could lead: based on early adopters, it suggests a team of two to five people could supervise an “agent factory” of 50 to 100 specialised agents running an end-to-end process such as onboarding a customer or closing the books. The human role shifts from execution to orchestration.
This creates a new kind of operating model. Responsibilities stay explicit, but tasks become dynamic. The organisation works less like a hierarchy of positions and more like a network of capabilities, where work flows to whoever, human or machine, is best placed to do it.
What replaces the job description
If the job description goes, something has to take its place, or accountability goes with it. The most practical replacement is an outcome charter: a short document for each team, or each person in a fluid team, that is reviewed every quarter rather than every few years. It covers six things.
- The outcome. What the team exists to deliver, in a sentence.
- The measures. How success is judged, in business terms rather than activity.
- Decision rights. What the team can decide alone, and what needs escalation.
- The AI systems it supervises. Which agents and tools the team is accountable for, including their quality and risk.
- Escalation paths. Who to go to when an AI system misbehaves or a decision exceeds the team's authority.
- Capabilities. The skills the team needs now, which feed hiring, development and internal moves.
The charter is lighter than a job description, but more explicit about the things that matter in an AI-native organisation: outcomes, authority and oversight. How the same shift changes individual responsibilities is explored in From Roles to Responsibilities.
The new architecture of work
Several other patterns are emerging in organisations experimenting with AI-native operating models.
Some are building internal AI platforms that provide shared capabilities across teams: knowledge retrieval, analytics, automation and decision support. Others are using internal talent marketplaces that let people move between projects based on their skills rather than their titles. Some are structuring teams around outcomes rather than functions, with AI embedded directly in everyday work.
The common principle is that work is defined by problems to be solved, not positions to be filled. The organisation chart becomes less rigid, and capabilities become the real currency of work.
What it means for careers
For individuals, the change alters how a career feels. Traditionally, a career followed a predictable path: enter a role, master it and progress through more senior versions of it.
When AI keeps changing the nature of work, roles start to look like temporary configurations rather than permanent professional identities.
Professional value lies less in performing specific tasks and more in applying judgement, insight and oversight within complex systems. In a single week, you might supervise automated processes, interpret what AI tools have produced and guide a team through an ambiguous strategic decision.
You are defined less by a job description and more by the outcomes you help deliver.
That can feel unsettling at first. It also opens possibilities. People are less confined to narrow functional silos, and skills can move more freely across projects, products and teams.
The career ladder starts to look like a lattice: a network of experiences rather than a single upward path. The stability of the old system may fade, but so do many of its constraints.
Accountability in a fluid system
Losing the job description does not mean losing accountability. If anything, accountability matters more.
When tasks move dynamically between people and AI agents, organisations have to be explicit about who owns outcomes, who supervises systems and who intervenes when problems arise. McKinsey's agentic-organisation research makes the same point: oversight has to become real-time and embedded, with people holding final accountability.
Responsibilities are defined around outcomes and decision rights rather than task lists. Escalation paths are clear, oversight is built into workflows, and people remain accountable for the decisions that matter.
The system becomes fluid, but responsibility stays visible.
Designing organisations without job descriptions
The job description is one of the oldest artefacts of the modern organisation. For more than a century it has helped companies manage complexity by defining who did what.
AI changes the nature of the work itself.
Tasks move between people and machines. Processes adapt continuously, and teams reorganise around outcomes rather than functions. Static role definitions struggle to keep pace with how work actually happens.
The organisations that thrive will not try to preserve the old structure. They will redesign work: replacing rigid job architectures with outcome charters and flexible responsibility frameworks, and organising talent around capabilities rather than titles.
The traditional job description may not survive the shift. The organisation may become far more adaptable because of it.
