Certainly! Here’s the translation:
86% of Companies Have Increased AI Use in Their Security Teams, But Many Still Don’t Understand How It Works
On June 3, 2025, cybersecurity company Deep Instinct released the sixth edition of its report Voice of SecOps, focusing on the real impact of artificial intelligence (AI) on enterprise security operations (SecOps). While the use of AI has become widespread, the study reveals a more complex situation: a landscape of more sophisticated threats and a significant knowledge gap among professionals.
According to the report titled Cybersecurity & AI: Promises, Pitfalls – and Prevention Paradise, 72% of organizations have modified their cybersecurity strategies in the past year due to the rise of AI. Additionally, 86% have increased its integration into operational tasks. However, nearly two-thirds of respondents claim to have difficulties understanding the basic concepts of AI, and 38% cannot distinguish between machine learning and deep learning.
More Complex Cyberattacks and Environmental Pressure
One of the most alarming data points is the increase in AI-powered threats. 46% of organizations have detected a rise in targeted phishing attacks, and 43% have experienced impersonation attempts using deepfakes. Both cloud and on-premises storage are under pressure, identified as one of the main risk vectors by 83% of professionals, just behind phishing (84%).
In response, 82% of organizations have opted for preventive strategies, and 64% have faced direct pressure from the C-suite to adopt more proactive defensive measures.
Operational Benefits, But at Human Cost
The report also highlights that while AI has reduced the time spent on manual tasks—an average of 12 fewer hours per week per team—70% of professionals believe its integration contributes to workplace burnout. This is compounded by regulatory complexity: 32% report difficulties keeping up with new regulations on AI, and 37% fear that these rules will lead to short-term economic penalties.
Moving from Reaction to Prevention with Deep AI
In light of this scenario, Deep Instinct proposes a new approach: preventive data security based on deep learning, an advanced form of AI. Its DSX (Data Security X) platform anticipates even unknown threats with a detection rate over 99% and false positives below 0.1%. The system can block threats before they execute in less than 20 milliseconds.
“Detecting and responding is no longer enough. It’s reactive, costly, and can’t compete against AI-enhanced threats,” says Lane Bess, CEO of Deep Instinct. “Only with a truly preventive approach can we regain control.”
The company has also incorporated GenAI into its system with the DIANNA tool, capable of providing explanations for unknown threats in under 10 seconds.
Conclusion: Much Work to Be Done
The report reflects an industry in transition: with increasing adoption of artificial intelligence but without a full understanding of its capabilities and risks. The preventive approach emerges as a necessity rather than an option in an environment where cyberattacks are becoming increasingly sophisticated.
The study is based on a survey conducted by Sapio Research with 500 cybersecurity leaders in U.S. companies with over 1,000 employees, spanning critical sectors such as healthcare, infrastructure, technology, energy, and public administration.
Source and report at Deepinstinct