AI Talks #5 with Adam Skiles, Chief Information Officer at Huntington University
Every campus wants to “do something with AI” right now. Far fewer have asked the uncomfortable question underneath the excitement: is our data even ready for it?
That is where this episode of AI Talks lands. Host Sathish sat down with Adam Skiles, CIO of Huntington University, who has spent twelve to thirteen years building IT services in higher education and brings a K-12 background. Adam is the person who has to make AI work inside a real institution, with real budgets, faculty, and risk. What follows is a conversation about why most universities are not as AI-ready as they think, why the answer starts with data, and how a mission-driven institution decides where AI belongs. We have kept it close to how the two of them talked, with the key findings after each exchange.
Adam Skiles
Adam Skiles is Assistant Vice President for Information Technology and CIO at Huntington University, a Christian liberal arts institution. He leads enterprise IT strategy, digital modernization, and cybersecurity, building secure, scalable digital experiences across complex higher-education ecosystems.
Sathish Kumar Mariappan
Sathish Kumar Mariappan is Co-Founder of BinaryWorks, an Atlanta-based certified Drupal agency. He helps education, government, healthcare, and enterprise organizations with Drupal development, migration, AI consulting, and digital strategy, building secure, scalable digital experiences across complex ecosystems.
Episode TL;DR
- → AI’s biggest blocker is not the model. It is the data. “If the data’s bad, the results you’re going to get are going to coincide with that.” Clean data, modern systems, and governance come first.
- → Higher ed and industry are finally in the same boat with AI, opening two-way partnerships where universities supply AI-skilled talent, not just receive internships.
- → Faculty fall into four camps: active integrators, curious experimenters, formation-first skeptics, and principled refusers. You have to bring all four along.
- → Mission beats hype. Every AI use case at Huntington is tested against two questions: does it align with the mission, and does it serve human formation?
- → Money is real but available. Indiana’s Lilly Endowment has put roughly $500 million behind AI readiness; Huntington is pursuing about $5 million over three years.
- → Skills are shifting from output to understanding: students who can explain, critique, and defend what the tool produced, not just prompt it.
- → Governance is non-negotiable: explainability, accountability with a named AI system owner, bias detection, and validity checks.
- → People come before tools. Change management and an AI Fellows program matter more than any platform; Huntington runs AI through BoodleBox (an aggregator) under a tool-agnostic responsible-use policy.
- → The pain points ahead: sustainability after grant money runs out, and cognitive overload for students and staff.
- → The prediction: over the next three to five years, the line between higher education and industry will blur dramatically.
From K-12 to CIO: What Higher Ed Is Actually Preparing Students For
Sathish: “You’ve been director of IT services at Huntington for twelve, thirteen years, and you also worked in K-12. What drew you to higher-education technology, and did that K-12 background help?”
Adam: “Yes, it has. Education is a very different sector from many others, but K-12 and higher ed are similar in a lot of ways. What we’re trying to do is meet students where they are, allow them to grow, find their calling, and pursue that wholeheartedly. Whether that’s K-12 or higher ed, that mission is still very central.”
Sathish: “As a CIO, what challenges do you face with so much changing around us?”
Adam: “Probably the biggest challenge is that our environment is very much changing. The workforce is changing. Some of the programs we’re running now, we weren’t running ten years ago. And that pace is changing.” Many of the careers his students will fill, he added, “are probably not even there yet.” So the job is not only to teach a specific skill. It is to teach students “how to adapt and be agile” for roles that do not exist yet.
Adam frames higher ed less as a content-delivery business and more as an adaptability business, which is why he treats AI as a long-term capability question rather than a feature to bolt on. For institutions weighing AI adoption in higher education, the starting insight is strategic, not technical: the value of AI is not in chasing the newest tool, it is in building students and staff who keep adapting.
Prefer to Listen? Stream the Full Episode on Spotify
A summary can only carry so much. The full conversation has the pauses, asides, and reasoning that make Adam’s perspective land. It is an easy listen for your commute.
🎵 Listen on Spotify →Why Higher Ed and Industry Are Suddenly in the Same Boat
Sathish: “With AI gaining momentum after ChatGPT took off, what opportunities do you see for higher education?”
Adam: “What we’re finding is that many organizations, especially small to medium-sized businesses, are really in the same space we are when it comes to AI: how we’re adapting to it, how we’re bringing it to our employees, our customers, our students.”
Adam: “The workplace has usually been a little ahead of where education is, or education has sometimes been a little ahead. The fact that we’re at the same place, in both our knowledge and how we’re using and moving forward with it, is very different. It allows for partnerships we’re not used to.”
The dynamic is flipping. Universities normally host student internships at local businesses; now, Adam said, “those organizations are looking to us to provide them a resource in students who possibly have skill sets they don’t already have.”
AI has briefly leveled the playing field. Neither side has it figured out, and that symmetry creates real two-way value: students gain experience while local employers gain AI-literate talent. The underrated opportunity in AI strategy for higher education is positioning the institution as a regional AI talent hub while the academia-industry gap is unusually small.
The Four Kinds of Faculty: How Educators Really Feel About AI
Sathish: “How’s adoption going with your workforce? Are they okay with it, or do they see it as a threat?”
Adam: “That’s a great question. I’d say it’s mixed, not really pro-AI or anti-AI.” From focus groups and conversations with their own faculty, four distinct groups emerged.
| Faculty Type | What It Looks Like |
|---|---|
| Active integrators | Already using generative AI, building use cases, doing lesson planning with it. |
| Curious experimenters | Tinkering with consumer tools, but unsure where they’re permitted to use them. |
| Formation-first skeptics | Deeply concerned about critical thinking and writing development. Not anti-AI, but unwilling to compromise on core student outcomes. |
| Principled refusers | View generative AI as a net negative for student learning and formation. |
Adam: “Our employees on the operational side span the same spectrum, but those individuals are working with more sensitive data, and therefore they’re a little more trepidatious about the risk that comes along with these tools.”
Adam’s refusal to flatten people into “adopters versus resisters” is the point. A skeptic worried about student writing is not an obstacle; they are a quality-control signal. For leaders managing faculty AI adoption, the lesson is that change management has to be segmented: the message that wins over a curious experimenter will not reassure a formation-first skeptic.
Where AI Helps, and Where Human Formation Has to Win
Sathish: “Can student experience be improved significantly with AI, or do you think that won’t be the case?”
Adam: “Many institutions are going to have to decide where we use it and where we don’t. Where does it make sense and where does it not?” For Huntington, two foundational principles settle most arguments: any AI use “must align with the mission of our institution, and it must serve human formation.”
Adam: “Does this AI use truly allow us to form the whole person? Or does it detract? Where does human agency and human intervention need to be paramount and not lost?”
Sathish: “You connect with a lot of higher-ed CIOs. What concerns or opportunities keep coming back?”
Adam: “A lot of use right now is on the operational side. Many existing tools already have AI infused in them, so we evaluate whether that helps or detracts.” From there, some institutions move to specific academic use cases in curriculum and assessment. A recurring goal is freeing people up.
Adam: “Where can we help our faculty support our students in ways where right now they’re bogged down with administrative work that isn’t enhancing the student experience, so faculty can spend more time on the activities they don’t currently have time for?”
Adam is not asking “what can AI do?” but “what should AI do here, given who we are?” That mission-and-formation filter keeps the institution from adopting AI just because it is available. The strongest guardrail for responsible AI adoption is not software; it is a principle applied consistently.
“Your Data Is Broken”: Why Data Readiness Comes Before AI
Sathish: “Students now start their college search on ChatGPT, not a brochure. On enrollment, is there a strong need for tools there, and what are you working on?”
Adam: “The core of any AI use, as you move forward, is the data systems you have. If the data’s bad, the results you’re going to get are going to coincide with that. You can’t just use a tool with bad data, unclean data, or data that isn’t governed well.”
Huntington built a dedicated data system modernization team focused on clean, accurate, well-governed data. Then came the anchoring line.
Adam: “The core of any use of any AI tool is the data that sits behind it. If the data’s bad, the results you’re going to get are going to coincide with that. You can’t just use a tool with bad data, unclean data, or data that isn’t governed well.”
Sathish: “Many institutions run outdated ERPs that aren’t in the cloud and can’t connect to AI. How can those institutions scale?”
Adam: “Your core data systems are really where you have to start. If they’re not where they need to be, implementing an AI tool in that environment is going to be less productive and probably more costly, and it’ll introduce additional risks around transparency, data privacy, and security.” His own Huntington playbook ran in order: infrastructure first (hardware, the data center), then modern CRMs and a new ERP on top. “Getting those core infrastructure pieces in place creates a launching-off point where you can innovate more readily, more easily, and more securely and reliably.”
Sathish: “So it depends on the use case, build or buy?”
Adam: “Right. It’s not going to be one size fits all.” Many existing tools already have AI built in, including their analytics platform’s natural-language reporting.
“Your data is broken” is not an insult. It is a diagnosis. AI sits on top of data the way a building sits on a foundation, and Adam’s sequencing (infrastructure, then data, then governance, then AI) explains why some schools get fast wins while others light money on fire. For any organization planning AI adoption in higher education, data readiness, data quality, and data governance are not back-office chores for later. They are the precondition for AI returning value instead of risk. A tool layered on poorly governed data amplifies privacy and security risk. That groundwork, the data modernization and governance Binary Works focuses on, is what has to happen before you scale the AI.
Watch the Full Conversation on YouTube
Reading the highlights is one thing; watching Adam reason through these decisions in real time is another. The full episode includes the back-and-forth and the examples that do not fully translate to text.
Who Pays? The Lilly Endowment and the Cost of Getting AI-Ready
Sathish: “All this data modernization and cleanup carries significant cost. How are institutions budgeting for it?”
Adam: “You’re right, it’s a significant spend.” Indiana institutions have an unusual tailwind. The Lilly Endowment, he explained, has “infused half a billion dollars over the next two to three years in the state of Indiana to help higher-education institutions prepare their faculty and their students.”
Adam: “We’re looking to receive probably in the neighborhood of about five million dollars over the next three years, if everything goes well.” Initial planning-grant funding already helped them develop their AI strategy and the guiding principles that govern their use cases, and assess where their students, employees, and region stand. His instinct, again, was partnership: “What’s going to be key is not just institutions doing this on their own, but partnerships.”
Even with half a billion dollars in play, Adam spent the planning grant on strategy and principles before tools, the discipline most rollouts skip. Grant funding for AI in higher education is substantial, but the institutions positioned to win it can articulate a coherent AI strategy, a readiness assessment, and guiding principles.
Teaching Students to Think, Not Just Prompt
Sathish: “Entry-level jobs are getting replaced by AI. How are you preparing students for a market that wants different skills?”
Adam: “There’s been this gap between competency and understanding. People are using tools without fully understanding the underlying pieces that produced the outcome.”
His fix is to change what gets assessed, grading the thinking rather than just the output.
Adam: We’re not just assessing them on their outputs. We’re helping students understand their critical thinking, what goes into it, and how to present that and speak to it when those things are questioned.” In the real world, someone will eventually push back: “Tell me about the results. Why are you seeing it this way? You have to understand what went into it, what the core data is.”
Adam: “Those skills aren’t tool-specific. Explainability, transparency, awareness of safety and security: they’re general skills that help students adapt as the tools change.”
Adam is redrawing the line between cheating and learning. If the deliverable is all that gets graded, AI breaks the system; if the reasoning gets graded, AI becomes a tool students still have to understand. For educators rethinking assessment in the age of AI, the actionable core is simple: design assessments that require students to explain, defend, and critique their work, not just produce it.
Online Learning, Faster Degrees, and the New Credentialing
Sathish: “More students want online education over coming to campus. Do you see that, and are you preparing for that shift?”
Adam: “Many institutions aren’t necessarily looking at doing an AI program. They’re looking at how AI gets infused into existing things, how we further enhance existing programs and make them more readily accessible.”
The pace is accelerating. Some four-year programs have already compressed to three, and Adam expects that to speed up. Credentialing is shifting too.
Adam: “Are students wanting to wait for a four-year degree to say they’re credentialed? There’s going to have to be credentialing throughout the process, showing ‘I’ve gained this knowledge, I’ve gained this skill,’ and being able to present that to employers.”
Adam’s vision is not a flashy standalone “AI degree.” It is AI woven into existing programs to make them faster, more flexible, and more verifiable, with stackable credentials that match how students and employers now think about proof of skill. The keyword is infusion, not replacement: use AI to accelerate delivery, broaden access, and credential learning continuously.
Catching Students Before They Fall: AI and Retention
Sathish: “On student retention, are you using any tools to predict when a student might be at risk of falling out?”
Adam: “Yes, we’re implementing a retention suite. We’re doing regular pulse surveys, gauging students’ level of engagement, how they’re feeling, how they think they’re doing, any emotional stressors that come into play.” The aim is to surface support at the right moment by pulling signals from the learning management system and other engagement tools, “so that we’re not too far along when the student needs the resource before they receive it, or before they even know a resource is available to them.”
This is AI at its most humane: not automating students away, but noticing the quiet ones before they slip through. Predictive retention is one of the clearest, lowest-controversy AI use cases in higher education, and it depends entirely on the foundation Adam stressed earlier: clean, connected, well-governed data.
Governing AI: Vendor Risk, Data Governance, and Non-Negotiables
Sathish: “You’re on Huntington’s risk management committee. How do you handle risk and data governance with so many third-party tools?”
Adam: “Vendor risk management is a big thing when it comes to AI, because there are so many tools.” His solution leans collaborative. Through the Independent Colleges of Indiana, schools that share many of the same vendors pool their assessments: “Rather than each of us doing vendor risk on the same vendor twenty different times, let’s do it once and share that information.”
Sathish: “And data governance? With hallucination, a tool can sometimes surface data to everyone. Who should have access?”
Adam pointed to a set of explicit principles.
| Principle | What It Requires |
|---|---|
| Explainability and transparency | AI use is disclosed to stakeholders; outputs and AI-informed decisions are understandable and documented. |
| Accountability | A named AI system owner takes responsibility for results and ongoing value; users ensure the work reflects actual understanding and intent. |
| Fairness and bias detection | Bias is assessed before implementation; fairness risks are documented and reviewed regularly. |
| Validity and reliability | System owners regularly review accuracy and relevance of both the inputs and the outputs. |
Each principle in Adam’s governance assigns a human a job. The “AI system owner” idea stands out: someone is accountable not just for turning a tool on, but for proving it keeps delivering trustworthy value. For any institution drafting an AI governance framework or responsible AI policy, these four pillars (explainability, accountability, fairness, validity) are a strong, reusable template.
Strategy Before Tools: Where Institutions Should Start
Sathish: “What’s your advice for someone thinking about implementing AI at their institution?”
Adam: “The first thing they have to set aside is: what is your strategy overall as an organization with AI? And that can’t be a purely IT exercise. It has to be an institutional exercise with buy-in.”
That means structure. Adam is working with Dr. Becky Benjamin, a professor of psychology, to lead a new group, and standing up an Office of Academic Effectiveness that unifies previously disparate pieces: faith integration and vocation, the center for teaching and learning, and academic technology.
Adam: An AI strategy isn’t just an IT strategy. It sits between organizational strategy and IT strategy. You need a good cross-functional group across the institution to decide on your strategic AI principles.”
Those principles become guardrails, the non-negotiables. “Humans must remain accountable for all meaningful decisions, and truthfulness and integrity determine all of our AI use.” Get those in place, Adam argued, and you can decide faster, building the “innovation engine” that lets the rest take hold.
Sathish: “So is a top-down approach best?”
Adam: “I don’t know that it’s necessarily top-down. In a university you’ve got healthcare experts, agriculture experts; you need them to feed up. The innovation has to come from the bottom up. But as an organization, we have to decide the parameters: where does this fall in, does it align with our mission?”
Adam’s model is neither top-down nor bottom-up; it is both, with clear roles. Leadership sets the guardrails and the mission filter; the people closest to the work generate the use cases. This is the part most “AI rollout” advice gets wrong: a successful AI strategy for higher education is cross-functional and principle-led, not an IT project with a budget line. Define your non-negotiables first, then let innovation flow up from the experts on the ground.
Bringing People Along: Change Management, AI Fellows, and BoodleBox
Sathish: “Does the workforce need to be trained for these new things?”
Adam: “Human capacity and change management is probably the number one pillar for us. How do we bring our people forward? Because we can’t bring students forward where our people aren’t.”
Faculty used to being the expert in the room are now using tools they are not yet expert in, sometimes with students who know more. To bridge that, Huntington is building an AI Fellows program to bring staff and faculty along over several years, leaning on free and paid AI courses plus credentialing to show progress.
Sathish: “If your people are at different levels and see tools in the market, do you recommend they subscribe, or should it be reviewed at the organization level given the risks?”
Adam: “We’re implementing a product called BoodleBox. It’s kind of an AI aggregator. There’s not a single AI tool that’s perfect for every use case.” It brings ChatGPT, Claude, Gemini, and others into one platform, helps users pick the right tool, and is pre-built for higher ed with security and compliance built in: “a tool they can really start running with.”
Underneath it sits a deliberately tool-agnostic responsible AI use policy. “These are the guidelines if you’re using AI. These are the data pieces you can use with it and can’t. We’ve given them guardrails, what you can and can’t do, without saying ‘use this tool, don’t use that tool.’ We give them latitude, but with strict guidelines.”
Adam separates two things most institutions tangle together: the policy (tool-agnostic, durable) and the tools (changeable, evaluated case by case). The sequence is the lesson: invest in people first, then write a tool-agnostic responsible-use policy that outlives any single platform. Tools keep changing; a good policy and a confident workforce are what compound.
Staying Current and the Real Pain Points: Sustainability and Cognitive Overload
Sathish: “How do you keep up with so many changes and new launches every day?”
Adam: “You could spend every waking second just feeding yourself full of AI.” His answer is community over solo effort, leaning on consortiums like HESS (a higher-education systems and software consortium), the Independent Colleges of Indiana, the Council of Independent Colleges, and EDUCAUSE, while watching the major labs (Microsoft, Anthropic, OpenAI) directly. “Something often starts with a particular vendor or thought leader and then ends up becoming a core tenet across the space.”
Sathish: “What big pain points does higher ed still face that AI tools might solve in the future?”
Adam: Adam named two. First, sustainability: “Especially when you’re receiving grant funding, how do we keep that sustainable over time?” Second, and more human, cognitive overload: “We used to say, ‘here’s a new tool, great for this use case.’ Now we’ve got tools with a multitude of use cases, and it’s almost an overload of ‘I don’t even know where to start.'” The job, he said, is not just teaching the tool. “It’s showing them how to handle that cognitive overload, the stress and emotional impact, so they can say, ‘I’m competent in what it’s producing and I can speak to it.”
Adam treats “keeping up” as a filtered team sport: consortiums do the triage so he is not blindsided. And he puts the human cost of AI front and center. The bottleneck is not capability anymore, it is capacity. Helping people build confidence is becoming as important as any technical rollout, in education and industry alike.
A 3-to-5-Year Prediction: When Industry and Higher Ed Blur Together
Sathish: “Final question: what’s your prediction for higher education over the next two to three years?”
Adam: “Five years is probably a bit long. If we’ve seen how the pace has changed in the last two years alone, it’s been exponential.” His bet for the next three to five years is a surge in partnership between industry and higher education.
“I see this back-and-forth, higher ed helping industry, industry helping higher ed, not only improving how we operate, but how we educate our students and the opportunities we provide them. That line between education and industry is going to continue to blur.”
He pointed to companies building their own internal universities and expects that to grow, especially among small and medium-sized businesses.
Sathish: “Any final word for the audience thinking about AI?”
Adam: Whatever you do with AI, stay true and core to your mission as an organization. Set those guiding principles in place out of the gate, so it doesn’t compromise who you are. Higher education’s real business is forming humans. It’s a critical age and time for our students, and it’s key that we don’t lose that.”
The conversation closes a loop. Adam opened by calling education an adaptability business and ends by insisting that adaptability never override mission. The future he predicts, education and industry merging, only works if institutions keep human formation at the center: move fast on the tools, stay anchored on the why.
FAQ: AI Readiness in Higher Education
Why aren’t most universities ready for AI?
Because the foundation comes first. As Adam Skiles puts it, “the core of any use of any AI tool is the data that sits behind it. If the data’s bad, the results you’re going to get are going to coincide with that.” Most institutions have outdated data systems, unclean data, or weak data governance, so AI layered on top underperforms and adds privacy and security risk.
What should come first, data or AI tools?
Data and infrastructure. The recommended sequence is infrastructure, then clean and governed data, then AI tools. Implementing AI before fixing core data systems is, in Skiles’s words, “less productive and probably more costly.”
How should a university start adopting AI?
With strategy, not software. An AI strategy “can’t be a purely IT exercise.” It sits between organizational and IT strategy and needs a cross-functional group to set guiding principles and non-negotiables — humans accountable for meaningful decisions; truthfulness and integrity governing all AI use — before any tool is chosen.
What are the core principles of responsible AI in higher education?
Four pillars: explainability and transparency, accountability with a named AI system owner, fairness and bias detection, and validity and reliability. Every AI use is also tested against two questions: does it align with the mission, and does it serve human formation?
How is AI changing assessment and student skills?
The focus shifts from output to understanding. Instead of grading what students produce, institutions are redesigning assessments so students can explain, defend, and critique their work, building durable skills like critical thinking and explainability that transfer across tools.
Is Your Institution’s Data Ready for AI?
If this episode hit a nerve — aging systems, messy or ungoverned data, no clear AI strategy, faculty spread across all four adoption camps — you are not behind. You are where most institutions are. The difference is what you do next.
That gap between “we want to use AI” and “our data and governance can support it” is where Binary Works helps higher-education and organizational teams. From data readiness and modernization to a practical, mission-aligned AI strategy, we help you build the foundation before you scale the tools.
- • Data readiness assessments and governance frameworks tailored to higher education
- • AI-enhanced digital platforms built for institutional security and compliance
- • Mission-aligned AI strategy grounded in your institution’s guiding principles
Let’s talk about where your data and AI roadmap stand today →
Keep the Conversation Going
This is AI Talks #5 by BinaryWorks, inside stories and strategies from leaders navigating the shift to AI, automation, and the digital infrastructure they require.
Also Watch: AI Talks #4 Your Boat Is Leaking: How AI Exposes the Hidden Revenue Crisis in Higher Ed →