Artificial intelligence is quickly becoming a core feature in corporate cyber-defense. Yet many companies still struggle to move from ambition to real-world results for their Security Operations Center – SOC.
A new global study by cybersecurity firm Kaspersky shows that almost all organisations planning to build a Security Operations Center now view AI as essential. The survey found that 99% intend to integrate AI into their security processes. About 67% say they will probably adopt it, while 32% say they definitely will.
The appeal is clear. Companies expect AI to improve threat detection, speed up investigations and automate routine tasks around the SOC. 57% believe AI will help spot suspicious activity through automated data analysis. Another 49% expect it to enable faster, automated responses to incidents.
The main motivations are practical. Nearly half of respondents want better threat detection. About 44% aim to reduce manual work, and the same share hope to improve accuracy while cutting false alarms. Large enterprises are leading the push, with more ambitious plans to use AI across multiple security functions i.e SOC.
SOC Evolution: Closing the AI Execution Gap
But the study also highlights a wide gap between plans and execution.
The biggest obstacle is the lack of high-quality training data, cited by 37% of organisations. Without reliable data, AI models struggle to deliver accurate results. A shortage of skilled AI staff is another major barrier, reported by 32% of respondents.
Costs also remain a concern. About 31% point to the expense of building and maintaining AI-driven systems. The same share worry about new threats linked to AI use itself.
These challenges suggest that while AI is becoming a must-have in cybersecurity strategies, adoption will not be straightforward. Many organisations still need better data, skilled talent and clearer deployment models before AI can deliver on its promise inside the modern SOC.
