AI is increasingly used for repetitive IT tasks, but system administrators are still wary of its accuracy and potential security risks.
- 67% of sysadmins expect AI to automate vulnerability prioritization within two years.
- Accuracy and reliability (79%) are the top AI concerns for IT professionals.
- Troubleshooting is the most common AI use case but also its biggest failure point.
- 40% of sysadmins worry about falling behind peers in AI adoption.
What tasks are sysadmins most eager to automate with AI? The landscape of artificial intelligence in IT operations is shifting rapidly, with system administrators increasingly leveraging AI for high-volume, repetitive tasks. Troubleshooting and log analysis now see significant AI deployment, with 41% and 35% of sysadmins, respectively, reporting its use in these areas – a notable jump from 2024.
AI Poised to Tackle More Complex IT Jobs
Looking ahead to the next two years, sysadmins anticipate AI will automate a range of critical functions. Vulnerability prioritization and server monitoring are leading the charge, with 67% of respondents expecting AI to take on these responsibilities. Detecting and remediating incidents, alongside patch management, are close behind at 66% and 65% respectively.
Security controls and compliance analysis are also prime candidates for AI-driven automation (65%), with post-incident reviews (57%) and providing IT staff guidance and training (55%) following suit. Even troubleshooting, a current AI staple, is expected to see higher automation rates, with 55% of sysadmins predicting its expanded role.
Tasks Less Likely to See AI Automation
However, not all IT tasks are seen as ripe for AI takeover in the immediate future. Managing single sign-on (SSO) and passwords remains a human-centric domain for 48% of sysadmins. Similarly, administering user permissions (44%) and file management (38%) are tasks where human oversight is still largely preferred.
Fewer sysadmins believe AI will automate defining system usage policies (34%), providing first-level IT support to end users (30%), and installing/maintaining software (29%). Interestingly, troubleshooting (24%), post-incident reviews (23%), and IT staff guidance (23%) also appear on the “less likely to be automated” list, suggesting a nuanced view of AI’s capabilities.
Persistent AI Concerns
Despite growing adoption, significant concerns about AI persist among IT professionals. A substantial 40% of sysadmins worry about being outpaced by colleagues more skilled in AI, a slight decrease from 45% in 2024. The primary drivers behind this hesitation are clear:
- 79% express concerns about AI’s accuracy and reliability.
- 78% cite data privacy and security risks as major roadblocks.
- 60% fear losing control over automated AI actions.
As AI integration deepens, its limitations become more apparent. Troubleshooting, a frequent AI application, is also where AI struggles most, with 30% of sysadmins reporting failures in this area – nearly double the previous year. Furthermore, 20% of sysadmins noted that AI implementations have led to operational disruptions.
Other areas where AI is not meeting expectations include log analysis, with 12% of sysadmins reporting failures, and first-level IT support tasks, also at 12%. Even AI designed to support IT staff, like providing training guidance, is falling short for 10% of users.
