Artificial intelligence agents are rapidly becoming part of the cybersecurity toolkit, with some of the world’s strongest security teams using the technology to accelerate research, troubleshoot problems and solve complex challenges.
New research from cybersecurity readiness platform Hack The Box shows that AI agents are particularly common among high-performing cybersecurity teams.
The platform’s three-year benchmark data revealed that 68 per cent of the top 25 teams had at least one AI agent, despite AI agents making up only 2.7 per cent of all registered accounts.
The findings highlight the growing role of artificial intelligence in cybersecurity and suggest that experienced professionals are increasingly turning to AI as a tool to complement their existing technical skills.
AI-assisted cybersecurity teams record faster results
Hack The Box found that teams using AI assistance recorded substantially higher solve rates than teams that did not use AI.
Across all active teams included in the research, AI-augmented teams achieved a 3.2-times higher solve rate compared with teams operating without AI assistance.
The performance gap was smaller among the very best teams. Within the top five per cent, AI-assisted teams recorded a 1.69-times higher solve rate.
However, their biggest advantage was speed.
According to the research, AI-assisted teams among the highest-performing groups completed cybersecurity challenges three to four times faster than teams relying solely on human effort.
The results suggest that AI can help security professionals reduce the amount of time required for certain research, analysis and technical tasks.
AI does not automatically produce better results
Despite the speed advantage recorded by AI-assisted teams, the research showed that artificial intelligence does not necessarily outperform the strongest human-only teams.
The highest-performing human team successfully completed all 36 challenges, while the strongest AI-assisted team completed 32 challenges.
This indicates that speed and overall performance are not necessarily the same thing.
AI can help cybersecurity professionals move through certain tasks more quickly, but experienced human operators remain important when dealing with complicated problems that require judgement, strategy and verification.
Human expertise remains essential
Hack The Box researchers said AI appears to provide the greatest benefit when it is used by people who already understand the cybersecurity problems they are attempting to solve.
AI agents can assist with research, coding, troubleshooting and initial analysis. However, human professionals still need to decide which approaches to take and determine whether AI-generated information is accurate.
This is particularly important because AI systems can produce incorrect or unreliable information.
Rather than simply accepting an AI-generated answer, cybersecurity professionals need to understand the underlying problem well enough to question the output and verify the results.
The research therefore points towards a collaborative approach in which artificial intelligence supports skilled professionals rather than completely replacing them.
Cybersecurity teams are solving challenges faster
The Hack The Box research also identified a major improvement in cybersecurity challenge performance over the past three years.
The median time required to solve challenges fell from 26.1 hours in 2024 to 13.8 hours in 2026.
At the same time, more teams were able to complete the entire challenge board, despite the number of challenges increasing.
Only two teams completed the full board in 2024. The figure increased to three teams in 2025 before rising significantly to 15 teams in 2026.
Hack The Box said the improvement could be linked to several factors, including better preparation, reusable tools, increased familiarity with the platform and deeper specialist expertise.
The growing use of AI may also be contributing to the changing way cybersecurity professionals approach complex technical challenges.
AI is becoming part of cybersecurity operations
Haris Pylarinos, founder and CEO of Hack The Box, said the cybersecurity industry is increasingly moving beyond the question of whether AI will become part of security operations.
The bigger question, he suggested, is whether organisations can integrate AI into cybersecurity workflows safely and effectively.
The latest findings show that AI agents are appearing most frequently alongside highly skilled cybersecurity professionals.
Rather than replacing human expertise, the technology is being used to enhance existing capabilities by helping professionals complete certain tasks faster.
As AI agents continue to become more capable, cybersecurity teams are likely to place even greater emphasis on human judgement, technical knowledge and the ability to validate machine-generated results.
For organisations defending against increasingly sophisticated cyber threats, the emerging model could be a combination of AI-powered assistance and human-led decision-making.
The research ultimately suggests that the future of cybersecurity may not be about choosing between people and artificial intelligence, but about finding effective ways for both to work together.
WhatsApp Chat With Us Have a question? Send us a message. ›
