Data-first AI security strategy becomes critical in 2026

As AI adoption accelerates across APAC, cybersecurity leaders warn that governance frameworks are struggling to keep pace. As autonomous systems gain deeper enterprise access, a data-first AI security strategy is becoming essential. Andy Zollo of Thales has identified limited visibility into enterprise data and weak governance as the biggest risk for 2026. Meanwhile, seven in ten APAC organizations now cite AI as their top data security concern, while credential theft remains the leading cloud attack vector.

The rise of agentic AI intensifies the challenge. Unlike traditional tools, autonomous AI systems can authenticate, access enterprise data, and make decisions at machine speed. If organisations have not clearly defined what data AI systems can access, retain, or share, those decisions may be made implicitly through system behaviour, creating avoidable compliance and exposure risks.  

A strong AI security strategy begins with data classification. Organisations can secure AI workloads only if they know what data they hold and how sensitive it is. Identity governance is equally critical, with AI systems requiring least-privilege access controls, authentication oversight, and audit trails similar to human users. Encryption also remains a major weakness, with nearly half of sensitive cloud data in APAC still unencrypted.  

Agentic AI also raises the stakes for cyber defence. AI-powered attackers can automate credential theft, intelligence gathering, and adaptive attacks at scale. For enterprises, that means traditional perimeter-based security models are no longer sufficient. 

The strategic takeaway is clear: businesses must treat AI security as a data governance challenge, not simply a tooling problem. Organisations that establish visibility, identity control, encryption, and AI governance early will scale AI more safely and with greater confidence. Those who delay may face incidents that force action under far less favourable conditions. 

Source: 

https://www.itnews.asia/news/a-data-first-ai-strategy-is-critical-to-managing-security-threats-in-2026-625777