Fighting AI with AI: A cybersecurity blueprint for 2025

AI is both sword and shield in today's cyber war. From adaptive ransomware to deepfake phishing, threats are evolving—fast. But with AI-driven detection, Zero Trust, and human oversight, organizations can fight smarter and stay secure in 2025

- Ashok Pandey

In a world where cyber threats evolve as rapidly as the technologies built to stop them, the battlefield is shifting— and artificial intelligence is right at the heart of it. In this exclusive interview, we speak with Kamalasekar Subramaniam, Principal Consultant at VTRAC, Cybersecurity Consulting Services, Verizon Business, on how AI is reshaping both the offense and defense in cybersecurity, and what organizations must do to stay ahead in 2025.

AI in Cybersecurity: Double-Edged Sword

AI is certainly the talk of the town. On one side, organizations are using it to detect and respond to threats through technologies like SoC automation and anomaly detection. On the other, threat actors are leveraging AI to craft more sophisticated, targeted attacks. We're entering an era where the only way to fight AI is with AI.

Organizations need to shift to AI-based detection and response strategies. Behavioral detection, user analytics, deception technologies, and automated hunting are no longer optional—they’re essential.

Beyond Rules: Behavioral & Predictive Analytics

Traditional systems rely on rules. AI-based behavioral detection is different—it’s model-driven and trained on large datasets. It learns how users behave and flags even subtle deviations. Compared to conventional approaches, it’s far more accurate.

Pair that with predictive analytics-where behavior is combined with threat indicators— and you're moving towards proactive security. It’s this fusion that offers the most accurate form of anomaly detection today.

Automation with Purpose: Threat Hunting Gets Smarter