In AI diffusion, the lack of 'backup'... determines the data resilience for corporate survival

As artificial intelligence (AI) spreads, the ways data is generated, moved, and exposed are rapidly changing. Some analyses point out that today, a company’s competitiveness no longer relies solely on simple “backup,” but on “data resilience”—the ability to actually recover and maintain operations. This means that how quickly a system can be restored after a failure has become a core indicator of organizational resilience.

Krista Kess, Chief Analyst at TheCube Research, recently noted that companies no longer view data protection as a back-end task. As Software as a Service (SaaS), hybrid cloud, multi-cloud, and AI-based workloads increase simultaneously, data visibility decreases and control becomes more difficult. In this process, assessments suggest that Veeam Software Group is shifting from being a pure backup solution provider to integrating recovery, security, and governance into an operational layer.

Kess specifically explained that AI not only increases data volume but also creates new dependencies within models and entire pipelines, making recovery more complex. In fact, only 31% of organizations currently back up more than half of the data generated by AI. This indicates that, compared to the speed of AI introduction, protection policies and verification systems have not been consistently applied.

More important than “protection” is the actual possibility of recovery

Market discussions are also changing. In the past, the focus was on whether data was protected; now, it is more about demonstrating the feasibility of recovery in actual failure scenarios. Some argue that although many within organizations assume they are “ready,” this confidence often proves fragile when real incidents occur.

Kess believes the biggest current risk is the combination of “downtime” and “loss of control.” Because whether dealing with ransomware, complying with regulations, or maintaining customer trust, without recovery capability, each is difficult to defend. Especially, surveys show that many organizations have not fully achieved the expected results when using AI for data protection, revealing a gap between technological adoption and operational maturity.

Leading companies are treating recovery as an “implementable function,” rather than a passive security measure. They repeatedly conduct recovery drills under real conditions, incorporate recovery processes into incident response plans, and introduce automation to shorten the time from detection to action. This perspective views data resilience as part of enterprise risk management, linked to security, privacy protection, costs, and regulatory compliance.

Recovery in the AI era is aimed not at files, but at “business state”

Krista Kess, Chief Analyst at TheCube Research, commented that Veeam’s recent releases demonstrate the trend of market changes in the AI era. The key is that the concept of recovery needs to go beyond data-level rescue and evolve toward restoring the actual operational state of the enterprise.

He believes that in an AI environment, simply restoring files, workloads, and applications may no longer be sufficient. Because new operational states are intertwined with agent activities, policy controls, approval workflows, permissions, and contextual information. Ultimately, future recovery is likely to extend beyond data object restoration to restoring the “business state.”

This also means that the more actively a company adopts AI, the more it needs to redefine data resilience as a core pillar of business continuity, rather than just an IT function. In the market, besides recovery speed, the scope of recovery, verification systems, and whether it is linked to governance are becoming variables that determine competitiveness.

Focusing on the recovery strategy outlined at VeeamOn

SiliconANGLE Media’s live studio theCUBE will focus on data resilience as a main topic during the “VeeamOn” event on May 14. Industry leaders and practitioners will discuss how to build an operational model that goes beyond traditional backups, integrating recovery, security, and governance.

This discussion coincides with the rapid change in data protection concepts driven by AI proliferation. For enterprises, it is increasingly important to go beyond “storing” data and establish a recovery system that can sustain actual operations even in crises and be verifiable. Some analyses suggest that in the AI era, the key to success depends not only on the speed of technological adoption but also on how quickly systems can recover to normal when shaken.

TP AI Notice: This summary was generated using a language model based on TokenPost.ai. Its main content may have omissions or inaccuracies.