As cyber threats grow more sophisticated and cloud adoption accelerates, the fusion of artificial intelligence and cybersecurity is emerging as both a necessity and a competitive edge. In this exclusive conversation, Rajtilak Majumdar, Deputy CISO at PwC, unpacks how AI is transforming cloud security — from real-time threat detection and predictive analytics to governance frameworks, transparency, and even AI-versus-AI defence strategies. He also shares how this shift is redefining the role of the CISO in today’s hybrid, multi-cloud world.
AI’s role in real-time defence
PwC’s 2024 Global Digital Trust Insights revealed that one in four organisations suffered a cloud data breach in the past year. For Majumdar, AI’s greatest impact lies in automation and predictive analysis.
“Integrating AI with platforms like Microsoft Sentinel allows organisations to scan thousands of indicators of compromise within milliseconds — work that once took teams hours or even days,” he explains. “Similarly, AWS GuardDuty enables near-instant flagging of anomalies in vast activity logs. AI brings speed, scale, and proactivity — but human oversight remains crucial.”
Redefining the CISO’s mandate
With 53% of businesses already embedding AI into their cybersecurity strategies, the role of the CISO is no longer reactive.
“The convergence of AI and cybersecurity has transformed the CISO from a firefighter into a strategic risk manager,” Majumdar says. “Predictive tools and zero-trust frameworks empower CISOs to anticipate and neutralise threats, while also ensuring governance frameworks keep pace with evolving architectures and user behaviours.”
Closing the visibility gap in hybrid and multi-cloud setups
Cloud complexity and workload sprawl make visibility a top concern for security leaders. AI, Majumdar notes, helps bridge this gap.
“AI-driven platforms unify signals from endpoints, DLP, vulnerability management, and behavioural analytics. Instead of relying only on signature-based detection, they identify anomalies in user or application behaviour far more quickly, offering a panoramic view across users, applications, and infrastructure.”
The pillars of AI-enabled cloud security
Majumdar outlines six foundational pillars for real-time threat detection in the cloud:
- Zero Trust Architecture — continuous verification of every access request.
- Unified Data Collection & Integration — centralising logs and telemetry across the ecosystem.
- Advanced AI/ML Models — behavioural analytics and anomaly detection.
- SOAR platforms — automated responses to free up human expertise.
- Continuous Learning & Adaptation — evolving policies and processes to keep pace with adversaries.
- Cloud-native security features — leveraging built-in AWS, Azure, and GCP controls.
“Overlaying these with robust governance and threat intelligence ensures security that is both proactive and accountable,” he adds.
Balancing AI with explainability and governance
With regulators demanding transparency, explainable AI (XAI) is a growing priority.
“Advanced services exist, but whether they’re truly explainable remains a concern,” Majumdar admits. “Auditability is key — CISOs must insist on clear reasoning behind alerts, strong audit trails, and robust AI governance frameworks. Knowing where data resides and who accesses it is as important as the models themselves.”
Building AI-native defences against AI-powered threats
Adversaries are already weaponising AI — forcing CISOs to evolve threat models and adopt AI-native defences.
“AI-augmented threat hunting, dynamic deception techniques like adaptive honeypots, and AI-based red teaming are crucial,” Majumdar stresses. “But explainability and continuous monitoring are equally important. Zero trust, combined with AI-driven behavioural analytics, remains central.”
A race of AI vs AI
Asked whether the future of cybersecurity is a contest between “your AI” and “their AI,” Majumdar is unequivocal:
“Absolutely. Attackers know their tools and data intimately. Defenders must do the same — either building or rigorously vetting their AI. The future of cybersecurity will be a continuous race of innovation, adaptation, and resilience.”
