Exploring the Role Data and Trust in the Era ot AI – CIO Roundtable
By Lane F. Cooper, Editorial Director, BizTechReports and Moderator, CIO.com
While artificial intelligence (AI) holds the potential to unlock new levels of efficiency and innovation for businesses, sustained success will hinge on a wide spectrum of trust – from the underlying data to the community of professionals who act on the ensuing insights. Building trust in AI-generated insights will require both robust data management and human oversight.
Michael Brown — Iron Mountain
"Our roundtable session revealed that the role of data in the enterprise has evolved significantly from being a commodity, often siloed and critical but not strategic, to becoming a foundational element for IT and AI strategies," Brown said. "The potential of data has long been locked in repositories or applications, but there is growing recognition that AI is one of the mechanisms that can unleash it, allowing businesses to do more with their data."
The discussion, he added, explored how data, once seen as merely holding operational value, is now viewed as a strategic asset. Roundtable participants shared how, for many years, their businesses collected vast amounts of data but lacked the tools or vision to utilize them holistically.
“Data often remained fragmented across departments, limiting its integration and real-time use. However, the rise of AI has catalyzed a rethinking of data's role, transforming it into a critical driver for innovation and efficiency,” said Brown.
As a result, he observed, different industries are increasingly focused on ensuring the cleanliness and integrity of their data, particularly as they build AI programs.
"There was a clear consensus at the table around data ownership and quality being the key to successful AI implementation. Each organization must ensure the data they feed into AI systems is in the right format and comes from reliable sources," he added.
Trust vs Reliability
Bill Silvio — Iron Mountain
In the process, pointed out Silvio, roundtable participants described a significant shift in how their organizations view their data.
"For years, the cornerstone of IT management has been about reliability; it was about leveraging data insights to keep networks up and databases functioning," Silvio said. "Yet, our trust in the data itself has remained relatively low. I'm not sure we'll ever get to 100% trust in the data, but it's essential that we strive for it as AI becomes more embedded in enterprise operations."
One of the roundtable executives, recalled Silvio, provided an example from the healthcare sector to illustrate the complexities surrounding the trustworthiness of data in the context of AI-generated insights, describing a scenario where AI suggested a diagnosis based on medical data provided to the system.
The AI diagnosis, however, was overruled by a physician who was present in the exam room. The doctor noted information not on the medical chart, such as the fact that the patient was homeless and distressed, and used his human judgment to reach a different, more accurate conclusion.
"While the AI was technically correct in analyzing the data it received, the underlying issue lay in the source of the data itself—in this case, the patient. The information captured by AI was incomplete or misleading. It is a situation that underscores the importance of human oversight in AI-driven processes, particularly when it comes to critical decision-making," Silvio said.
Roundtable participants also noted a change in the process of data analysis. As businesses increasingly rely on AI, they are moving from a deterministic IT model, where outcomes are highly predictable, to a probabilistic one, where outcomes are more uncertain.
"It's not just about whether the data is reliable, but whether it's telling the truth. Consider, for instance, if a sensor generating data has an issue? Trust in AI will require a deeper understanding of the veracity of the data, not just its fidelity," he said.
The Human Element in AI Decision-Making
Brown reinforced the point, noting that AI is nowhere near being a "set it and forget it" system. Human involvement will continue to be essential, particularly when understanding the impact of AI on employees and end users. "Open communication with teams is critical to ensuring the successful adoption of AI," he said.
Looking ahead, both Brown and Silvio expressed optimism about the future of AI and its potential to reshape industries. Brown acknowledged that while skepticism around AI is understandable, trust in AI systems will grow as businesses see its value.
"It's going to be an exciting ride," he said. "AI will continue to transform how enterprises manage and utilize data. Fostering trust in these systems will be vital to unlocking their full potential," he concluded.
###