For the past two years, I have watched organizations rush toward artificial intelligence with the same urgency that companies once reserved for cloud computing and digital transformation. Boardrooms are demanding AI strategies. Investors are asking about AI roadmaps. Press releases announcing AI initiatives appear almost daily.
Yet from where I sit, many companies do not have an AI problem. They have a data quality problem.
That distinction matters because organizations are spending enormous amounts of money trying to unlock value from AI while overlooking the very thing that determines whether those investments will succeed. AI is only as good as the information it receives. If the data is incomplete, inaccurate, inconsistent, or misunderstood, the output will reflect those flaws at scale.
The technology is not the bottleneck. The data is. The enthusiasm surrounding AI is understandable. A survey shows that 88% of executives plan to increase AI-related budgets, while nearly 79% organizations report some level of AI agent adoption.
What concerns me is that many organizations are investing in AI before they fully understand the condition of the information that powers it. In many boardrooms, the sequence has become backwards. Leaders decide they need AI. Budgets are approved. Public announcements are made. Expectations are established. Only later does the organization begin to discover that the underlying data is fragmented, outdated, inconsistent, or simply wrong.
That is not an AI failure. It is a foundation failure.
The reputational risk may be even greater than the financial risk. When a CEO tells investors, directors, employees, and customers that AI will transform operations, those promises create expectations. If the technology fails to deliver because the underlying information was never reliable, the organization is left explaining why the results never materialized.
We are already seeing evidence that implementation is proving more difficult than many expected. Across industries, organizations are discovering that AI success depends less on the sophistication of the technology and more on the strength of the foundations supporting it. Research consistently shows that companies with stronger AI governance practices and more disciplined approaches to data management are more likely to achieve positive business outcomes from AI initiatives. That reality should not be surprising. AI can accelerate decision-making, automate processes, and uncover patterns at scale, but it cannot compensate for poor information. The quality of the output will always be limited by the quality of the data, processes, and oversight behind it.
One example that stands out to me is Apple. While competitors rushed to make increasingly aggressive AI announcements, Apple initially took a more measured approach. Investors questioned whether the company was falling behind. Yet Apple appeared willing to let others move first, observe what worked, and incorporate proven technologies rather than rushing to create everything internally.
There is a lesson in that strategy. Not every company needs to be first. Not every company needs to build everything from scratch. Sometimes leadership means resisting pressure to follow the crowd until the business has the right foundation in place.
I see a similar dynamic within the insurance industry. One of the most important datasets in commercial property insurance is a Statement of Values, often referred to as an SOV. It contains information about insured properties, including location, occupancy, construction characteristics, and value. Every underwriting decision depends on the accuracy of that information.
Recently, I reviewed a large portfolio where the results were eye-opening. Roughly half of the information associated with the properties was incorrect. Even more alarming, approximately 85% of the properties lacked accurate occupancy descriptions. In other words, insurers often did not fully know how the buildings they were insuring were being used.
When the analysis was completed, only about 1% of the properties contained the correct location, description, characteristics, and valuation within an acceptable range.
Now imagine building an AI system on top of that dataset. No matter how sophisticated the technology becomes, it cannot produce reliable insights from unreliable information. It will simply generate bad conclusions faster.
That is why I believe organizations need to stop treating AI as a technology project and start treating it as a business transformation initiative. The companies that create lasting value from AI will not necessarily have the most advanced models. They will have the most disciplined approach to information quality.
The good news is that perfection is not required. Many executives assume that if their data is not flawless, improvement is impossible. I disagree. If an organization has been operating profitably despite significant information gaps, moving from 1% confidence to 50% confidence can create tremendous value. Progress matters more than perfection.
What leaders need first is visibility. They need to understand what they know, what they do not know, and where the weaknesses exist. In my experience, knowing what you do not know is often more valuable than believing you know something that turns out to be wrong.
That process also requires expertise. Not one expert. Multiple experts. One of the biggest mistakes organizations make is assuming a single AI specialist can solve every challenge. AI is not a one-person project. It is not a silo. Successful implementation requires data specialists, business operators, subject matter experts, technology leaders, and executive decision-makers working together.
In fact, my advice to organizations evaluating their readiness for AI is simple. Get a second opinion. Then get a third. Different perspectives expose blind spots. Different experts identify different risks. The organizations that surround themselves with diverse expertise are far more likely to uncover the issues that could undermine future investments.
Ultimately, I believe the winners in the AI era will not be determined by who spends the most money or who makes the biggest announcements. They will be determined by who builds the strongest foundation.
The conversation around AI has become obsessed with algorithms, agents, and automation. Those technologies are important. But they are not where the battle for competitive advantage will be won.
The real question is much simpler. Can you trust the information feeding the system?
If the answer is no, the smartest AI in the world will not save you. If the answer is yes, AI becomes one of the most powerful business tools ever created.
The industry does not have an AI problem. It has a data quality problem. And the organizations that solve it first will be the ones that define the next decade of business leadership.
About the Author:
Todd Rissel is the co-founder and CEO of e2Value, a provider of property intelligence and valuation solutions for the insurance industry. With more than three decades of experience working at the intersection of data, risk, underwriting, and property valuation, he advises insurers and industry leaders on how data quality, emerging technologies, and market trends are reshaping decision-making across the property ecosystem.
