AI won’t kill entry-level cyber-security jobs, but it will redefine them  

It is not new for a major technological shift to trigger fears that it will eliminate entry-level roles and disrupt the talent pipeline. And AI is certainly no exception.  

 

Recent benchmark data from Hack The Box’s Augmented-AI vs Human Capture The Flag competition, the largest controlled comparison of AI-augmented and human-only cyber-security performance, found that AI-augmented teams achieved a 70% higher challenge solve rate than human-only teams. While elite AI teams with human-in-the-loop completed tasks faster, enabling up to 4.1x more output under pressure.  

 

The implication from this is clear. AI is already reshaping the early stages of cyber-security careers.  

 

The missing middle in cyber-security talent  

At first glance, AI appears to be a powerful leveller. Entry-level practitioners who are using AI tools can solve more problems than they could on their own. But this creates what could be described as a productivity illusion.   

 

AI can act as a competency bridge by helping junior analysts produce outputs that look comparable to those of more experienced practitioners. But this is without giving them the opportunity to develop the underlying skills they will need to validate, challenge or direct those outputs effectively.   

 

The tasks that have traditionally formed the foundation of cyber-security careers, such as triage, basic analysis and pattern recognition, are the ones that are most likely to be automated. In our benchmark data, AI showed much higher solve rates on very easy and easy challenges.  

 

This creates a structural problem for the industry. If AI absorbs the tasks that have been traditionally training new cyber-security professionals, how do we develop the next generation of expert talent? Without intervention now, we risk creating a missing middle. This means a gap between entry-level access and senior capability that could undermine long-term resilience.  

 

AI augments teams, it does not replace them  

Despite the potential disruption to career development, our benchmark data also paints an optimistic picture. AI does not eliminate the need for human expertise. It changes where that expertise matters most.    

 

At the elite level, the gap between AI and humans narrows significantly. The best human team in the benchmark solved 36 out of 36 challenges, compared with 32 out of 36 for the top AI-augmented team. This shows that creativity, reasoning and dynamic problem-solving remain important human strengths.  

 

The same data also shows that AI’s advantage peaks on medium-difficulty tasks and then narrows on harder ones, with some hard challenges exposing clear failure modes. Creative domains such as coding and reversing also show near-parity at the elite tier.  

 

This means that the strongest model is not simply AI adoption – it is AI used effectively within human-in-the-loop teams. The advantage does not come from AI alone. It comes from building AI-fluent teams that know how to direct it, validate it and apply judgement under pressure. The full benefit of AI is dependent on how well practitioners can direct agents, validate outputs and apply judgement on the hardest problems.  

 

Redesigning the cyber-security career model  

The challenge for security leaders is not around whether they should adopt AI, it is how to do it effectively without breaking the talent pipeline. This needs a fundamental rethink of how cyber-security careers are developed and structured.  

 

Firstly, entry-level roles must evolve. Instead of focusing on the repetitive tasks that AI can automate, they should be redesigned around AI-assisted workflows. Junior practitioners need to learn how to interpret outputs, validate results and understand failure modes.  

 

Secondly, organisations need to invest in dual skill development. Cyber-security fundamentals must be taught alongside AI orchestration skills. The ability to direct, prompt and govern AI systems effectively is quickly becoming a core competency.   

 

Thirdly, training environments need to simulate real-world complexity, so practitioners learn by doing, rather than by simply observing AI outputs. This means using AI-assisted training environments that focus on learning by doing and effectively rebuild the apprenticeship model.  

 

Finally, human mentorship becomes more important, not less. Experienced practitioners need to focus more on oversight, coaching and decision-making. They need to ensure that their knowledge continues to transfer even as workflows change. As part of this we need to build new mentorship paths that include AI governance skills.  

 

AI is a strategic necessity, not a talent crisis  

There is a tendency to automatically frame AI as a threat to cyber-security jobs. But that narrative misses the point. The real risk is not that AI will replace entry-level roles. It is that organisations will fail to adapt those roles and their training models in time.  

 

AI is already proving effective at increasing the speed and scale at which both defenders and adversaries operate. In our benchmark, the data shows that AI-augmented teams demonstrated consistent performance gains across all levels, with measurable efficiency improvements even outside elite tiers.   

 

This isn’t the first time we’ve seen a shift like this. When computers first reshaped the workplace, things kept evolving, and they will again. The organisations that succeed will be those that treat AI as more than a cost and efficiency tool. It should also be treated as a catalyst for redesigning how talent is developed, deployed, and retained. Cyber-security is a human challenge. AI will just reshape how we solve it.    

  

---

 

Haris Pylarinos is CEO & Founder at Hack The Box  

 

Main image courtesy of iStockPhoto.com and fotosipsak