The biggest obstacle to AI isn’t technology anymore. It’s uncertainty.
Every new AI breakthrough promises a competitive advantage. Every new platform claims to accelerate transformation. Yet beneath the excitement lies a far more practical question that every technology leader is trying to answer: Where do we begin without making the wrong investments?
Rob Wellen, Data and AI Partner at Aprio believes the answer isn’t found in chasing the latest AI innovation. It begins with building an AI readiness framework – one that aligns business strategy, trusted data, governance, and organizational priorities before AI becomes another disconnected initiative. In this conversation, he shares why the organizations creating lasting value are the ones treating AI as a business capability, not just a technology deployment.
Strategy and AI readiness framework
Before organizations scale AI, they must first build an AI readiness framework rooted in strategy, alignment, and measurable outcomes.
You often speak about “measurable transformation.” What does a successful AI transformation look like from a business outcome perspective?
Wellen: It looks like a few things. For some organizations, measurable transformation is super simple, just having your organization aligned and executing a common AI strategy. That is a successful transformation for a lot of organizations right now. It’s just being able to have consensus on what to do, what value AI can bring, and have agreement on where to make AI investments.
Second, is all around business value. The organizations that are winning right now have aligned and defined their strategy and tied the outcome or transformation to their business plan for revenue growth, cost reduction, faster cycle times, or better decisions, with evidence and value attached to each.
I have worked with organizations that are well past experimentation, where AI is driving real outcomes, and this is just how they operate. And the fastest movers all got one thing right: they slowed down before they built. They defined the business problem first, like any normal initiative. Not the tool, not the platform — the problem, and the value of solving it.
Many mid-market organizations feel pressure to adopt AI quickly. In your view, what should companies prioritize first: AI tools, data modernization, governance, or organizational readiness?
Wellen: If I had to pick, I would start with data modernization and governance as low-hanging fruit to get started now. Both are foundational elements of an effective AI readiness framework. It will enable your AI strategy for the long term, and CTOs can typically own those workloads. Data modernization to secure your data in a good format to use for AI and any kind of analytical reporting, machine learning cases, computer vision, or AI assistant development. And governance puts the right policies, organization, and technical controls in place to enable responsible AI across the organization. If people do not have trust and security in their data, it is very hard for organizations to adopt AI successfully.
How should CTOs evaluate whether their organization is truly “AI-ready” versus simply interested in AI?
Wellen: Many organizations and CTOs I talk to are interested in AI. But the real difference in being “AI ready” comes down to the organizations that have taken a business approach to it. They have defined their business problems, and they know what they need and want to do.
If I were a CTO today trying to determine whether I was AI-ready, I would evaluate three things.
- First, I would evaluate my current technology stack and talk to my vendors to confirm that I had the right technology partners in place to enable AI.
- Second, I would evaluate my data security protocols and set up basic security controls to prevent data loss and protect my data.
- And third, I would run a small AI planning workshop with my business leaders to identify and evaluate their interest in AI and determine real business needs.
From there, a CTO can pick up a low-risk, high-value starting point. He/she can develop a pilot solution, get the data into a good place, or build an AI roadmap. Any of those would move you forward, help determine if you are AI ready, and get your business ready to use AI at scale.
Every AI readiness framework I have helped organizations build over the last five years has been different. Because every business starts from a different place. Over the years, I do not think anyone ever said they were truly AI-ready.
Like many things in life, there is no perfect time or event that says you are “AI ready”. But sometimes you have to take the leap. AI is still moving super-fast. And the key thing is to do something now, not wait, so we can evolve with the changing market conditions.
What are the most common foundational weaknesses you see when companies begin building AI initiatives?
Wellen: I have noticed four constant, foundational weaknesses.
First, many organizations still do not have a data strategy that brings together a comprehensive view of their enterprise information in a secure, governed environment. Their data is siloed across different systems. There is no single version of corporate truth, and there is little strategy for how to use that data to enable AI. That makes it extremely difficult to build AI assistants and workflows at scale.
Second, several organizations do not have an internal business champion or a dedicated team driving AI forward. That leaves it entirely on the CTO to define the right AI strategy and enable the business alone.
Third, there is a general gap in AI education and understanding what AI transformation really is. Organizations can see AI as a tool rather than a business enabler that can help them scale and grow.
Finally, data security and privacy are top of mind for a lot of organizations right now. Yet many still under-invest in it. That erodes trust in AI and ultimately limits adoption.
Data infrastructure and AI foundations
AI is only as intelligent as the data that powers it. Before enterprises can scale AI, they must modernize their data ecosystems, strengthen governance, and build cloud-native foundations that support trusted decision-making.
Many organizations still struggle with fragmented data environments. What practical steps can technology leaders take to improve data quality and governance before layering AI on top?
Wellen: That is a great question, and one I get often. The CTOs who do this well tend to follow a simple path.
Start by establishing lightweight data governance. That means deciding who owns what data, business or IT, and who is responsible for its quality. A simple approach is: the IT manages the data platform, while business teams own data quality. After all, they understand how the data is used every day.
From there, put technical controls behind that governance, starting with a common data platform for your enterprise data. A unified platform keeps your data secure and gives every team access to the same trusted information. That’s essential for building a strong AI strategy.
Third, establish simple policies to manage data consistently. For example, create a data governance council that brings the right stakeholders together to make data decisions.
Fourth, use AI assistants to detect data quality issues and recommend improvements to master data. This will help strengthen the very data foundation that AI depends on.
How important is cloud-native architecture in enabling scalable AI workflows and automation?
Wellen: Cloud-native architecture is very important. This is where the future of AI is already headed. If you look across the technology landscape today, there is tremendous investments and R&D flowing into cloud, data, and AI platforms. Cloud-native is where the leading AI capabilities and models land first – the managed services, the compute, the constant stream of new models embedded.
You are building on a platform that improves beneath you. Rather than one you have to keep rebuilding yourself and staying on top of all the changes. Businesses should focus on tools where there is substantial funding, active research and development, and growing market adoption.
AI governance and responsible AI
AI governance is no longer just about compliance – it’s becoming a competitive advantage. Wellen explains why organizations that establish trust, accountability, and responsible AI practices early are better positioned to innovate at scale.
As AI adoption accelerates, how should enterprises balance innovation speed with governance, compliance, and risk management?
Wellen: That is one of the most important questions enterprises are wrestling with right now.
The organizations that move fastest with AI innovation are usually the ones that build in the right guardrails early. Rather than treating governance as a brake or an extra step in the process, they put a few things in place from the start. E.g., a responsible-use policy to guide their AI development, an AI council to manage adoption and risk, and basic data protection to monitor and prevent data loss. That is what lets them move quickly while keeping their risk in check.
They essentially slow down to speed up. I call this the “trust layer” for AI. And trust is not just a technical problem, it is a people one. When employees, customers, and stakeholders trust that your data and AI is secure, ethical, and responsibly governed, adoption accelerates and friction drops across the board. The enterprises that get this foundation right are the ones that scale AI successfully.
What does an effective AI governance framework look like in practice? Especially for organizations that may not have large, dedicated AI teams?
Wellen: There are three components that belong in any AI governance program: AI policies, AI governance teams, and AI governance management. These are just one part of your overall AI strategy and data governance strategy.
- First, AI policies. These include your responsible-use policy, your AI mission statement, and any other policies you need to govern AI use.
- Second, AI governance teams and the meetings that support them. Three groups are needed in nearly every case: an executive AI group that drives overall strategy and adoption, an AI community group that helps track how people innovate and helps with adoption, and a weekly project meeting to stay on top of in-flight AI initiatives.
- Third, AI governance management. This is a newer term. But it covers your ongoing AI operations and cost management, including token consumption, model management, and health. This tracks adoption and ROI realization over time. Together, these three components form part of the broader “trust layer” I mentioned.
The foundation that lets you scale AI with confidence rather than guesswork.
Are companies underestimating the operational and reputational risks associated with deploying generative AI too quickly?
Wellen: Yes, they are. Many organizations are quick to deploy a pilot for customer service or website leads, but they don’t check the data as they deploy. Although the traditional data science use cases like forecasting, predictive analytics, and customer segmentation are well understood to have established risk frameworks and organizations have years of experience managing them, generative AI is a different risk profile entirely. Many organizations are applying the same comfort level they have with traditional machine learning to a technology that behaves very differently and is still figuring out its best practices, security model, and support.
There is also a data privacy dimension that is underappreciated. Many teams are feeding sensitive client or employee data into commercial LLM APIs without a clear understanding of the retention and training policies of those providers. In regulated industries, that is a real compliance exposure, and organizations have not done the vendor due diligence to understand what happens to their data once it leaves their environment. That diligence must happen before deployment, not after an incident.
Leadership, talent, and organizational change
The organizations realizing the greatest value from AI are rethinking leadership, upskilling talent, and fostering a culture where business expertise and AI capabilities evolve together.
Successful AI adoption is often more about people and processes than technology. What cultural or leadership shifts are necessary for organizations to truly become AI-driven?
Wellen: As a leader, you must model and drive the change yourself. The biggest cultural shift is the move from a bottom-up, technical tools approach to a top-down, business-led one. This is where CEOs, CTOs, and senior leaders drive the change, keep the focus on business outcomes, and rethink how the operating model itself must evolve to become AI-driven. That leadership shift and mindset are essential and can be the catalyst for AI driving change in an organization.
How should CTOs rethink talent strategy and workforce enablement as AI becomes embedded into daily operations?
Wellen: The talent conversation has shifted from “hire data scientists” to “build AI fluency across the organization.” The most value comes from people with deep business-domain experience who are already in your organization and apply AI to their own workflows. People in finance, operations, sales, and HR learn to work effectively with AI tools and AI-generated outputs. That requires a different kind of investment and a different way of thinking about talent.
The first piece is how you upskill and enable your current team. This is about setting the right culture for learning, creating space for people to experiment with AI, grow their skills, and embed it into the way they work, rather than treating training as a one-time event.
The second piece is the profile to hire and develop for. One of the most valuable people in the market right now are those who can translate between a business problem and a technical solution, who can speak to the CFO and the data engineer on the same day, and own an AI product end to end.
And the third piece is that you do not need to have all the right people in place right now. There are emerging ways to close that gap. You can bring in the idea of a forward-deployed engineer (FDE) who embeds directly in your teams, and there is growing demand for outsources data and AI offices to drive the organization’s work around AI strategy, AI solutions, and AI data.
The future
In the future, technology alone will no longer create competitive differentiation. Long-term success will depend on building capabilities that competitors cannot easily replicate.
Looking ahead three to five years, what will separate organizations that genuinely become AI-native from those that simply use AI tools occasionally?
Wellen: The organizations that will differentiate over the long term are the ones that build their AI capabilities around their deep domain experience – what makes them special – instead of chasing every opportunity. AI capability, once embedded in operations, processes, and experience, is not easily replicated and creates differentiation in the market.
The window to build capability is still open, and organizations can either start now by building their AI strategy or start building AI or data tools now to drive that advantage and move their organizations forward.
Key takeaway:
The real competitive advantage in AI may not be intelligence at all – it may be discipline.
While many organizations race to deploy the latest AI capabilities, those creating lasting value are investing in something less visible but far more durable: an AI readiness framework that helps them decide where AI should be applied, how it should be governed, and why it matters to the business. As Rob says: technology will continue to evolve at breakneck speed, but organizations that build strong foundations today will be the ones shaping tomorrow’s competitive landscape.
This article was originally published on CTO Magazine on July 10, 2026.
