Presented by Splunk
Agentic AI is making IT and security teams dramatically more efficient. But it is also removing the apprenticeship that has long produced experienced operators.
As organizations automate more of the work done by junior analysts and engineers, they are facing a challenge that is as much about workforce design as architectural design: how to build the next generation of experts when AI takes over the work that once trained them.
what is the junior workforce doing
For two decades, the path to becoming a world-class SecOps analyst, SRE, or NetOps engineer was one of repetition.
Trying false positives. Search the dashboard for references. Reading the logs at 2am which turned out to be benign. The industry treated this work as hard work, and in many ways it was.
But it also served as an apprenticeship.
The thousands of hours an analyst spent monitoring traffic patterns built the intuition that made them invaluable when a real attack came. That intuition isn’t taught in any one curriculum or captured in any runbook. It was accumulated through exposure, pattern recognition, failure, and growth. Over time, people acquire deep analytical experience this way.
However, agentic AI is now beginning to automate tasks that once served as the training ground for that expertise. There’s no reason for it to be slow. Hard work was expensive. The burnout was real. Organizations should use agents to reduce labor wherever possible.
Also, as we remove that apprenticeship loop, we need to offer operators something better in its place. How organizations approach this issue today will determine the winners for the future.
Organizations that deliberately adopt this approach will prepare the skilled operators to succeed in the next decade. Organizations that insist on this may find themselves with faster systems today, but with fewer people who understand them deeply enough to rule them tomorrow.
When automation saps accountability
There is another dimension to this conversation that receives less attention than it should.
In a regulated environment, the rigors of apprenticeship are part of the accountability layer. Frameworks from SOX to PCI DSS to HIPAA to NIS2 recognize that a series of human decisions are behind a control decision.
Auditors do not interview models. They interview people who can explain why a system did what it did, why the decision was correct, and whether the correct controls were in place.
When the population of professionals explaining that series begins to decline, the risk may not be immediately apparent. Control may still pass. The workflow can still be executed. The dashboard may still appear green.
But the underlying organizational memory begins to hollow out.
This is not just a problem of tooling. It is also a workforce skills and design issue. And for organizations moving rapidly toward adopting agents, the risk is closer than many think.
Building human expertise to control AI
When we lose part of the accountability layer to agents, humans will step into a different type of governance role. Controlling an agentic system means implementing automatic guardrails that adapt to and ensure non-deterministic agent behavior.S Agents behave appropriately in circumstances that no one fully anticipated. This means creating growth criteria that capture the right anomalies without overwhelming humans with the wrong anomalies. This means implementing dynamic tools, alerts and processes to review machine decisions to detect drift, bias and logic failures that no individual case would uncover.
The ability to evaluate and respond to these exceptions requires judgment built from years of experience, recognizing the learning patterns that the old apprenticeship model produced.
That’s why the workforce question and the architecture question are now the same question. If we expect humans to control increasingly autonomous systems, we need intentional pathways that help people manage the scale and speed of AI systems while building the intuition and judgment into the human operators needed to perform that task.
In the AI age, the most valuable platforms will not automate most tasks. They will help people become more capable, more reliable and more essential as the systems around them become faster and more intelligent.
This means organizations need to invest in a full ecosystem of expertise for operators: communities that disseminate shared practices, certifications or other credentials that make expertise visible, and human-oriented clarification and verification in AI, as well as learning pathways that build capacity. Empowerment is an architectural design choice
Human empowerment is an important part of the conversation around the practical use of AI. However, without a deliberate strategy to support it, it risks becoming the kind of phrase that means nothing because it can mean anything.
Empowerment cannot be merely an ideological requirement for agentic systems. It should be a set of design choices based on the behavior of the system. An agent system that empowers its human operators and enhances their professional skills does four things:
1. Exposes the logic behind it with the data genealogy
Every recommendation made by an agent must be traced back to the data it considered, the logic it applied, and the origin of the inputs it used. Operators who can see the logic develop judgments about when to trust it. Directors do not just hand over findings.
2. Divide authority with confidence and influence
Familiar, low-risk patterns can be handled autonomously. Novel situations or activities with a significant blast radius should proceed by default. The limit should be clear and configurable by the teams who own the results.
3. Treats disagreements as signs of improvement
When an experienced engineer ignores an agent, they are doing more than disagreeing. They’re perfecting the system with a decision that wasn’t in the model: a delicate dependency, a quirk in the environment, a constraint the data never saw. A system that registers an override but ignores the logic behind it learns nothing from a moment when a human knew better.
4. Captures resolutions as cross-domain knowledge
How an incident unravels is a lesson that rarely stays in one lane. A SecOps incident could expose an ITOps weakness. Network problems may impact business. When that connection remains only inside a closed ticket, the next team to hit it starts from zero. Resolutions should go to different areas and not end up where they were filed.
These are not aspirational qualities. Those are testable product capabilities. Leaders evaluating agentic systems must be able to identify where these capabilities reside, what happens when they fail, and whether operator skills improve after deployment.
The next advantage that comes when humans and AI scale together
For AI systems to be practical, reliable, and work at scale, the key design point is that AI work closely with and empower human operators.
Thus, the Agentic Age is not a story about replacing humans. It is the story of redesigning the systems that humans operate so that these operations can occur at machine speed and scale, while also increasing human expertise. Not at the expense of each other, but together.
That result is not a given. This will only happen where leaders take operator growth as a priority, not an afterthought. To achieve this, agentic systems must be intentionally designed to uncover logic, capture learning, and send work back to humans in a way that builds skills and careers, not destroys both.
Agents will become smarter and faster. The ability of the operators working with them to learn and grow in lockstep will determine whether the next decade of digital flexibility is something that organizations truly own, or something they rent from a shrinking pool of expertise.
Learn more about how Cisco Data Fabric powered by Splunk Platform Teams are being helped to accelerate agent operations.
Kamal Hathi is the SVP and GM of Splunk, a Cisco company.
Sponsored articles are content produced by a company that is either paying for the post or that has a business relationship with VentureBeat, and they are always clearly marked. Contact for more information sales@venturebeat.com.
<a href