AI Is Rewriting Work at the Task Level. We've Been Defining Work That Way for Nearly a Decade.

I recently read Anthropic's Scenarios for Our Economic Future, and one idea in it stopped me.

My immediate reaction was that we have been saying this for nearly a decade.

Long before AI agents became part of the workforce conversation, CyberSN was focused on understanding work at the task and project level. The reason was simple. If you do not understand the actual work being performed, almost everything downstream of it breaks.

You cannot bring in the right talent. You cannot accurately manage workload. You cannot see where one person is carrying too much responsibility. You cannot identify capability gaps. You cannot build meaningful training and development plans. And you cannot give people the professional efficacy that comes from knowing what they own, what is expected of them, and how their work contributes to the organization.

Eventually, a poor understanding of workload shows up as something every organization recognizes: burnout, turnover, execution gaps, and work that simply does not get done.

We have been saying this for years. AI has now made it impossible to ignore.

I have mixed feelings about that. I do not want the future of AI to be defined by how many jobs companies can eliminate. I care deeply about people having meaningful work, sustainable workloads, opportunities to grow, and careers where they feel effective and valued.

But AI is here. So the opportunity in front of us is to use it thoughtfully, and that starts by understanding the work.

Jobs Are Containers. Work Is Tasks and Projects.

Nearly a decade ago, CyberSN started building our cybersecurity job taxonomy because job titles were not giving us enough information. Two people could hold exactly the same title and spend their days doing completely different things.

So instead of focusing on titles, we focused on the work underneath them: tasks, projects, responsibilities, and capabilities.

What is this person actually responsible for? What work are they performing? What projects are they supporting? What capabilities does the organization depend on them to deliver?

We built our cybersecurity taxonomy around that level of detail nearly a decade ago. Almost three years ago, we expanded that thinking across IT.

At the time, our goal was not to prepare for an AI-agent workforce. We were solving a human and operational problem. Organizations could not effectively organize, develop, retain, or plan for talent if they did not have a common language for the actual work.

That was true ten years ago. AI has simply made it exponentially more important.

AI Is Changing the Work Inside the Job

This is what I found so interesting about Anthropic's scenarios. Their model looks beneath the job title and considers what happens to the tasks inside the job. Some tasks may remain largely unchanged. Some may be augmented by AI. Some may become automated. And entirely new tasks may emerge.

That is a much more useful way to think about what is happening inside organizations right now.

A cybersecurity engineer may still be a cybersecurity engineer six months from now, and their workload could look completely different. Perhaps AI handles pieces of investigation, documentation, analysis, reporting, or repetitive configuration work.

That does not automatically mean the engineer is no longer needed. It may mean:

• They finally have capacity to take on projects that have been sitting untouched.
• They can spend more time on higher-risk or higher-value work.
• The organization can address capabilities that have been understaffed for years.
• The employee has room to learn something new and expand professionally.
• Work that has been causing burnout can be redistributed.
• Entirely new responsibilities emerge around validating, governing, directing, and overseeing AI-generated work.

Workload Has Always Been a Human Issue

This is where I think something important is getting lost in the AI conversation. Understanding tasks and projects is not only about productivity. It is about people.

When organizations cannot see workload at the task level, they often do not know who is overloaded until that person burns out or leaves. They may not realize that one employee is the only person capable of performing a critical function. They may bring someone in based on a title and discover that the person's actual experience does not match the work that needs to be done. They may send employees to generic training rather than developing the capabilities the organization truly needs. They may believe they have enough people because the headcount looks right, while critical projects remain untouched.

That is why CyberSN has spent so many years focused on the work itself. Understanding tasks and projects allows organizations to manage both operational execution and the human experience of work.

You can make better hiring decisions. You can distribute workload more intelligently. You can create relevant development plans. You can identify people ready to take on more. You can see burnout risk earlier. You can improve professional efficacy. You can reduce unwanted turnover. And you can make sure the work the business depends on is actually getting done.

AI does not replace the need for any of that. It increases it.

AI Should Start With Workload, Not Headcount

This is one of the most important distinctions organizations can make as they adopt AI. If the starting point is a question about how many people AI can replace, I think you are beginning in the wrong place.

Start here instead:

1. What work needs to get done?
2. What tasks and projects make up that work?
3. Who is doing that work today, and what does their workload look like?
4. What is the best resource to execute each piece of work going forward?

That last question is where the options open up. It might be:

• A permanent employee.
• A contractor.
• A consultant.
• An MSP.
• An AI agent.
• A person using an AI agent and becoming dramatically more productive.

The objective should be making sure the right work gets done, workloads are sustainable, capabilities are covered, people have opportunities to grow, and the organization can execute.

That is a very different conversation from simply reducing headcount.

This Is Where Workforce Intelligence Becomes Critical

Most organizations already struggle to answer some very basic questions about their workforce. Add AI agents into that environment and the questions do not just multiply, they become harder to put off.

Every one of those answers has a shelf life. That is why I believe Workforce Intelligence has to become continuous. The work is going to keep changing.

What Will Your AI Agents Be Doing Six Months From Now?

This is the question I think every technology and security leader needs to start asking. Not simply what AI can do today, but what it will likely be able to do next.

AI capabilities are changing incredibly quickly. If an AI agent can handle several tasks today and twice that number six months from now, workforce planning cannot remain static.

That change affects workload. It affects hiring, contracting, budgets, training, and career paths. It affects burnout. It affects how teams are structured. And it affects what work your people should be spending their time on.

This is why workforce data at the task and project level becomes so valuable.

The Workforce Is Becoming an Ecosystem

At CyberSN, we already define workforce much more broadly than permanent employees. The workforce includes employees, contractors, consultants, MSPs, and the other resources organizations use to get work done. Now AI agents are becoming another part of that ecosystem.

That does not mean every resource is interchangeable. Far from it. It means leaders need a way to continuously understand the work, the capabilities required to perform it, the resources currently responsible for it, and the best resource mix going forward.

And throughout all of this, we cannot lose sight of the humans. Are workloads sustainable? Are people doing work that matches their capabilities? Are they getting opportunities to develop? Do they have the resources they need to be effective? Are we using AI to remove low-value work and create capacity, or simply adding AI on top of already overloaded teams?

Those questions matter just as much as what the technology can do. That is a level of Workforce Intelligence an org chart will never provide, and a job title will never provide it either.

Nearly a Decade of Defining Work Has Brought Us Here

CyberSN did not build taxonomies around tasks and projects because we knew exactly what AI would look like in 2026. We built them because work has always mattered more than titles.

We understood that if organizations could see the actual tasks, projects, responsibilities, and capabilities across their workforce, they could make better talent decisions, manage workload better, develop people better, reduce burnout, retain talent, and execute more effectively.

We have spent nearly a decade defining cybersecurity work at that level, and almost three years doing the same across IT.

Now the world is moving toward a workforce where humans and AI will increasingly execute work alongside one another. That makes understanding the work itself more important than ever.

For me, that is the opportunity in what comes next. Not replacing jobs. Understanding workload.

Understanding what needs to get done, and what is not getting done. Understanding how tasks and projects are changing. Understanding what humans should own, and where AI can create capacity. Understanding how that capacity can be used to make people and organizations more effective. Understanding where people need to grow, and where burnout risk is building. And continuously adjusting as the work changes.

AI is rewriting work at the task level. CyberSN has been defining work that way for nearly a decade.

That puts us in a strong position to help organizations understand not only what their workforce is doing today, but what their workforce needs to become next.

---

About CyberSN Workforce Intelligence

CyberSN's Workforce Intelligence Engagement gives cybersecurity and IT leaders a living view of the workforce ecosystem: the tasks and projects being performed, who and what is performing them, where capacity and capability sit today, and what work should move to an employee, a contractor, a consultant, an MSP, or an AI agent.

It is built on the same task and project level foundation described above, and it measures capability coverage, capacity, maturity, redundancy, and dependencies rather than headcount. So when the work changes, and it will keep changing, leadership can actually see what changed and decide what to do about it.