Build. Borrow. Buy. Bot.
AI agents introduce a fourth workforce strategy question: should this work be delegated to a bot?
Rethinking Workforce Strategy in the Age of AI Agents
For decades, workforce strategy has revolved around three fundamental questions:
- Do we build the capability internally?
- Do we borrow the capability through contractors or consultants?
- Do we buy the capability through acquisition, outsourcing, or technology?
Those questions still matter.
But they are no longer sufficient.
AI has introduced an entirely new workforce option that most organizations are still treating as a technology experiment rather than an operating model.
The fourth question leaders should now ask is simple:
Should a bot do this work?
Not every task belongs with an AI agent.
But increasingly, some work is practically asking for one.
The organizations that learn to distinguish between human work and agent work will create significant operating leverage over the next decade.
A New Workforce Category
Historically, organizations viewed work through one lens:
Someone had to do it.
If the work existed, a human was ultimately responsible for completing it, even if software supported them along the way.
Agentic AI changes that assumption.
Today’s AI agents are capable of executing repeatable workflows, synthesizing information, making routine decisions within defined guardrails, coordinating across systems, and escalating exceptions when judgment is required.
In other words, some work no longer requires a person to perform every step.
It requires a person to design, govern, and improve the workflow.
That is a fundamentally different operating model.
Not Every Process Should Become a Bot
This isn’t an argument to automate everything.
Far from it.
Organizations often swing between two extremes:
“We’ll automate everything.”
or
“AI can never replace our people.”
Neither is true.
The better question is:
What characteristics make work suitable for an AI agent?
Agentic workflows are particularly effective when work is:
- High volume
- Repeatable
- Rules-based
- Time sensitive
- Information intensive
- Cross-functional
- Dependent on pulling data from multiple sources
- Prone to delays caused by handoffs rather than expertise
Examples include:
- Preparing first drafts of executive reporting
- Coordinating interview scheduling
- Performing policy lookups
- Vendor onboarding
- Contract intake
- Compliance documentation
- Financial reconciliations
- Customer support triage
- Meeting preparation
- Knowledge retrieval
These are not necessarily strategic activities.
But they consume enormous organizational capacity.
Where Humans Still Win
AI agents are incredibly capable.
They are not executives.
Humans remain essential for work involving:
- Ambiguous judgment
- Ethical decision-making
- Complex negotiations
- Coaching
- Relationship building
- Creative problem solving
- Organizational leadership
- Accountability
The goal isn’t replacing people.
The goal is elevating people.
Every hour a leader spends gathering information instead of making decisions is an hour of strategic capacity lost.
Bots should absorb operational friction so people can focus on higher-value work.
Think of Bots as Digital Employees
One mental model I find useful is this:
Treat every AI agent like a member of your workforce.
Ask the same questions you would ask before hiring someone.
- What job are they performing?
- What authority do they have?
- Who supervises their work?
- How is quality measured?
- What happens when they make a mistake?
- How do they improve over time?
When organizations begin treating AI agents as governed members of the operating model, not just software, the conversation shifts from experimentation to execution.
The New Workforce Allocation Decision
For years, workforce planning asked:
Should we build, borrow, or buy?
Tomorrow’s organizations will ask:
- Should we build?
- Should we borrow?
- Should we buy?
- Or should we delegate this work to an AI agent?
That doesn’t eliminate the need for talented people.
It changes how we deploy them.
The future of workforce strategy isn’t simply about reducing headcount.
It’s about allocating work to the resource best equipped to perform it.
Sometimes that’s an employee.
Sometimes it’s a consultant.
Sometimes it’s software.
Increasingly, it will be an AI agent.
The leaders who master that allocation problem won’t simply become more efficient.
They’ll redesign how work gets done.