By Matthew Hardman
Asia is leading the world in AI adoption, but many businesses are beginning to realise that its effectiveness depends on how well data is structured, secured, and made accessible.
According to Hitachi Vantara’s latest State of Data Infrastructure Survey, 42% of enterprises in Asia now consider AI critical to their operations, outpacing the global average of 37%. Yet, this momentum faces a serious challenge.
The research found that AI models in Asia produce accurate outputs only 32% of the time, largely due to inconsistent data reliability. In many cases, information is scattered, unstructured, or difficult to retrieve, with businesses reporting that data is available when needed only 34% of the time. This fragmentation limits AI’s ability to deliver real-time insights and undermines its full potential. At the same time, data security concerns continue to grow, with 44% of businesses in the region ranking it among their top challenges, surpassing the global average of 38%. These figures highlight an uncomfortable truth: while AI is advancing, the underlying data infrastructure remains a major weak spot.
The 2024 global IT outage highlighted just how vulnerable enterprises can be when their backup and recovery strategies fail to keep pace with modern demands. Many had backup systems in place but still experienced extended downtime because their recovery plans were outdated, slow, or untested. With data volumes in Asia set to grow by 123% in the next two years, relying on traditional backup methods is no longer viable. The risks are not just about data loss but business continuity itself.
AI adoption often begins with proofs of concept and pilot projects, where organizations focus on innovation but may not fully consider long-term enterprise integration. Over time, these projects must transition into core business systems, requiring more than just functionality; they need enterprise-grade capabilities like resiliency, data protection, and regulatory compliance. Have businesses factored this into their AI roadmaps? Should organizations start adopting a model where every AI project includes an “enterprise tax”—a built-in budget allocation to cover the infrastructure and safeguards required to support AI at scale?
AI also presents new security challenges. Cybercriminals are now actively targeting AI training datasets and backup repositories, injecting manipulated data to degrade decision-making. Ransomware attacks have become more sophisticated, focusing not just on encrypting files but also wiping out backup copies entirely, leaving organizations with no path to recovery. Meanwhile, the growing complexity of hybrid and multi-cloud environments makes cohesive and scalable recovery strategies more essential than ever.
This World Backup Day, enterprises need to rethink their approach to data resilience. AI-powered backup integrity will become crucial in detecting anomalies before they spread, while immutable storage and zero-trust security models will help safeguard data from tampering. Resiliency planning must also account for the unpredictable whether it’s a cyberattack, system misconfiguration, or simple human error. Data observability, which uses AI-driven monitoring to ensure data remains compliant, secure, and instantly recoverable, must also be prioritized. With businesses increasingly operating across hybrid and multi-cloud ecosystems, seamless failover strategies will be essential to minimize downtime and maintain business continuity.
Asia’s rapid AI adoption is reshaping industries, but without trustworthy, high-quality, and resilient data, its full potential will remain unrealized. World Backup Day 2025 is not just about having a backup; it is about ensuring data remains secure, accessible, and instantly recoverable. AI and digital transformation demand a smarter approach to backup and recovery. The businesses that move beyond traditional backups to a true resilience strategy will be best positioned for the future.
The author is the Chief Technology Officer for Hitachi Vantara- APAC