Pavel Bykov is the Co-Founder and CEO of IP Fabric, which has been pioneering automated network assurance since 2016.
getty
For the first few years of enterprise AI adoption, most organizations were chasing the most cutting-edge capabilities. What models were performing the best? Which providers were innovating the fastest? How quickly would they be able to deploy AI at scale? But now, as AI becomes more deeply embedded in business operations, a more fundamental question has emerged: If the AI model you’re using isn’t really yours, then how can you trust it with your data?
According to the Linux Foundation, nearly four out of five organizations now consider data sovereignty a strategic priority. IT leaders want to know where AI is running, what data it’s accessing and, most importantly, who is controlling the underlying infrastructure.
AI Has Changed The Economics Of Enterprise Data
Twenty years ago, British mathematician Clive Humby coined the phrase “data is the new oil.” What he meant was that both data and oil need to be processed and refined before they can be used to their full potential. In 2026, I’d say a more appropriate saying is “data is the new solar energy.” Oil requires extensive preparation to be usable, and once it’s used up, it has no further value. Solar energy is also an upfront investment, but once solar panels are up and running, they keep producing power indefinitely.
Like solar panels, AI can reuse the same proprietary data again and again, increasing its value over time. For example, a single support transcript can be used to train an AI assistant, identify product issues and train new employees even months later. Organizations are realizing that in the age of AI, their data has near-endless potential, and IT leaders are seeking ways to establish stricter governance over how it is shared and used across their infrastructure.
Trust Is Reshaping AI Deployment
Many IT leaders draw parallels between AI adoption and the mass migration to the cloud. It took years for cloud providers to build up enough trust for organizations to feel confident using the cloud to store their data and run their critical workflows. But today, AI is moving faster than the trust-building process can.
While organizations want access to increasingly powerful AI capabilities, many are reaching the inevitable conclusion that having true ownership over their proprietary data far outweighs the benefits of using the newest or largest-hosted model. This is especially true for organizations that operate across critical infrastructure, who must abide by strict security and regulatory compliance standards. But even these strict standards aren’t enough to establish data sovereignty.
AI Is An Uncertainty Accelerator
AI is evolving faster than governance frameworks and regulatory environments can keep up. We’re still figuring out what “good” AI governance actually looks like, and the goalposts are always moving. Decisions that feel low-risk today may be intolerable in just a few years.
Some regulatory frameworks have introduced controls for AI, but if an organization is striving for sovereignty, it must turn the “check-box” mentality of compliance on its head, treating standards like GDPR and DORA as the bare minimum for data governance. Organizations must create more stringent controls on their own, and to do that, they need to understand where their data is and how it flows across every part of their infrastructure. Digital twin technology can give organizations the visibility they need to prove that they’re continuously in control of their data, even as technology, regulations and business priorities continue to evolve.
What IT Leaders Should Do Next
Organizations don’t need to fully own every technology they use; cloud, SaaS and managed services will remain essential parts of modern enterprise IT. What organizations do need, however, is the ability to govern where critical workloads run and how their data is shared.
Before implementing any third-party AI solution, organizations should pause to consider the following:
Are you protecting your data as a proprietary asset? Enterprise data is becoming increasingly valuable because AI can reuse it across countless future applications. Organizations should strive to establish controls beyond today’s regulatory requirements to effectively govern, access and reuse their own data.
Is your network optimized for flexibility? The AI landscape is evolving too quickly to assume that any one provider, model or deployment strategy will remain the best choice for long. Forward-thinking organizations are beginning to air-gap their architectures, thereby preserving flexibility as new technologies and regulations unfold.
Do you understand where your data is flowing and how it’s being used? True flexibility and control are both predicated on an organization’s ability to see every part of their infrastructure, from the dependencies that support critical business processes to the ways that their AI applications communicate.
It all boils down to this: The most innovative AI models today may not be the most trustworthy ones tomorrow. If organizations want to adapt to AI without sacrificing their data, the only long-term solution is to establish strict governance. And for that governance to be effective, it needs to be grounded in a complete understanding of your network infrastructure.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
>> Home