The AI Quandry: Why People, Not Technology, Are the Biggest Challenge to AI Adoption

ere is no question that AI is one of the most important global paradigm shifts we have seen in this decade. We have lived through the computer, the internet, smartphones, cloud services and social media. But AI feels uniquely different because of its ability to recognize patterns across enormous amounts of information, synthesize what it finds and help us make connections faster. It can make us dramatically more effective, accelerating research, analysis and decision-making in ways that would have taken far longer just a few years ago.

Last week, I attended the Digital Marketing World Forum at the Javits Center in Manhattan (#DMWF). I sat in the audience listening to panels and leaders from some of the biggest and brightest brands, Morgan Stanley, Chubb, Campbell’s and others, and one thing struck me. AI is already being used across marketing, sales, human resources, legal and procurement, but the governance and structure around that use still appear to be catching up. In many organizations there is an expectation of transparency, tell us when you are using AI, but not necessarily a consistent enterprise-wide framework for how it should be used. And that is where “shadow AI” begins to creep in: employees and teams using AI tools outside established or approved processes.

See content credentials

Shadow AI: When employees use AI and do not disclose

What surprised me is that this conversation is not new. More than four years ago, I attended another seminar hosted by a New York law firm with a major financial-services company, and AI adoption and governance were already on the agenda. So why, several years later, are so many organizations still wrestling with the same questions?

The latest McKinsey research makes the disconnect particularly interesting. Nearly nine in ten respondents say their organizations regularly use AI in at least one business function, and 44% say AI is now scaling across the enterprise. Yet only 37% report that AI has contributed positively to enterprise EBIT, (Earning before Interest and Taxes) essentially unchanged from last year. At the individual level, the story looks much stronger: 80% say AI has improved their productivity and 50% say it helps them make better decisions.

That gap is the quandary for me. We clearly know how to use AI. The harder question is whether we know how to change an organization around it.

McKinsey’s small group of AI high performers offers an important clue. These organizations are not simply dropping AI into existing processes. They are redesigning workflows around what AI now makes possible, while backing those changes with leadership commitment and operational discipline. In other words, the advantage is not just the technology. It is the organization’s ability to adapt around it.

So why are leaders getting stuck?

Everyone agrees AI is here, yet many leaders still seem psychologically paralyzed by its arrival at the doorstep. I believe much of the challenge is human, not technical. The technology is moving faster than leadership habits, organizational structures and cultures can absorb it.

The human barriers I see most often are:

• Poor management skills or unclear ownership

• Resistance and fear

• Old leadership paradigms

• Poor change management

• Organizational silos

• Unrealistic expectations

• Ethical, privacy or bias concerns

• Lack of alignment among key stakeholders

Mindset Matters in adopting AI

Overcoming the Scarcity vs. Abundance Mentality

AI adoption is becoming essential to competitiveness, but fear is a poor strategy for adoption. The longer an organization waits to understand where AI fits, the harder it becomes to build the internal muscle to use it responsibly and effectively. That does not mean adopting every tool or chasing every trend. It means learning how to adapt, experiment and make informed choices.

First and foremost, technology should enable humans—not to simply replace them. The opportunity is to remove low-value work, expand capacity, support creativity and give people better information with which to make decisions. In some cases that will change roles. In others, it may create entirely new ones. The leadership challenge is to help people see possibility rather than only scarcity.

What can a small or mid-sized company actually do?

This is where I think smaller companies can have an advantage. You do not need an enterprise-sized AI program to begin. You need a baseline, a few clear priorities and a willingness to learn. My firm works with a partner that provides a readiness framework reinforces this point: AI readiness is not a technical audit alone. It spans strategy and value, data, technology, talent and change, governance and risk, and the operating model around all of it.

The biggest chasm for many smaller firms is moving from scattered experimentation to a repeatable way of working. A focused 90-day period is enough to establish a baseline, identify gaps and begin building the foundation for scale.

1. Diagnose your culture and current use. Interview key stakeholders. Where are people already using Copilot, Gemini, Claude, ChatGPT or other tools? Where is AI helping, and where is it creating risk?

2. Assess readiness and alignment. Look beyond the technology. Do you have the right data, skills, ownership, approval paths and leadership support?

3. Choose a few high-value use cases. Start narrow. Prioritize opportunities based on business value, feasibility, risk and whether people will actually adopt them.

4. Design a human-first AI roadmap. Build AI into the workflow rather than bolting it onto an old process. Define what success looks like and who owns it.

5. Put responsible guardrails in place. Clarify approved tools, acceptable and prohibited uses, human oversight, privacy and security expectations, and escalation paths when something goes wrong.

This is a bite size approach that is realistic for most firms. Note, a lower readiness score is not a failure. It tells you where the first investment can create the most leverage. That is a much healthier way to think about AI maturity than pretending every organization should already be at scale.

The real inflection point

We are spending a great deal of time talking about what AI can do. I think the next conversation needs to be about what organizations, and the people inside them, need to become in order to use it well.

The winners will not necessarily be the companies with the most AI tools. They will be the ones that create enough trust for people to experiment, enough discipline to know where the boundaries are, and enough leadership courage to redesign how work gets done.

That is the real quandary of AI: the technology may be moving at extraordinary speed, but transformation still moves at the speed of people.

Is your organization in an AI quandry, if so, leave a comment on what is the biggest challenge your facing?

Sources:

Sources: McKinsey & Company, The State of AI in 2026: On the Road to ROI (August 2026); AI Readiness Checklist (2026).