AI models are advancing at an unprecedented pace, yet many large organizations remain frozen in place, stuck behind outdated IT policies, risk-averse leadership, and a fundamental literacy gap at the top. The technology is ready. The organizations, in many cases, are not.
Wharton professor Ethan Mollick posted a striking observation on X that captured one of the strangest dynamics in AI right now. While AI capabilities are advancing faster than at any point in the technology's history, there is a massive and growing adoption divide across the corporate world.
The contrast is stark. "It is one of the weirdest divides," Mollick wrote. "I speak to two companies in the exact same industry, and one has been using AI for the past 18 months. The other has a committee that has to approve every use case individually and talk about how AI companies will train on their data."
Mollick went further, pointing out: "It is amazing how many companies I've talked to still have AI effectively blocked by IT and legal departments for out of date reasons when many companies in highly regulated industries have figured out ways to deploy Enterprise ChatGPT, Claude and Gemini without any apparent problem."
The deciding factor, according to Mollick, is whether an executive is willing to assume risk.
This observation comes at the same time the Wall Street Journal reported that AI needs management consultants after all in a piece exploring why OpenAI announced its Frontier Alliance and Anthropic is signing deals with major consulting firms. The labs have built the technology. Now they need people with trusted enterprise relationships to actually get it deployed inside organizations that don't know where to start.
I discussed all of this with SmarterX and Marketing AI Institute founder and CEO Paul Roetzer on Episode 201 of The Artificial Intelligence Show.
None of this surprised Roetzer. In March 2023, two weeks before the release of GPT-4, he published what he called the "Law of Uneven AI Distribution." Its core thesis: "The value you gain from AI and how quickly and consistently that value is realized is directly proportional to your understanding of, access to, and acceptance of the technology."
He warned that "the impact and benefits of AI will be unevenly distributed to individuals, companies, and industries. In some cases, it will be by your own choice, and in others it will be by institutional design."
Three years later, those same three principles still hold. Understanding, access, and acceptance remain the primary blockers for most enterprises. The technology has transformed. The organizational barriers have not.
"In the last two months alone, I've spent time with leaders of major financial and healthcare institutions that are still blocking access to generative AI platforms in their companies."
—Paul Roetzer, founder and CEO of SmarterX
School administrators at both the high school and college level are still searching for answers on how to handle AI in the classroom. The key is to start with the fundamentals, Roetzer says, or you'll get nowhere.
"We're just going to be in this continual cycle of, 'Here we are three years later,' and those same three principles of lack of understanding, lack of access, lack of willingness to give up or accept the risks, are still preventing companies from doing anything with AI," Roetzer says.
The acceptance barrier is universal. Even Roetzer, one of the most AI-forward leaders in business, hasn't explored tools such as Open Claw or Claude Cowork yet because of security concerns and unknown risk. The acceptance principle applies at every level. The difference is in degree.
The literacy gap compounds the problem. Many companies have started bottom-up AI training programs, but Roetzer points out that this approach alone isn't enough.
"It has to also happen from a leadership level," Roetzer says, "because they won't give the priority and urgency needed to AI transformation if they don't understand AI capabilities themselves."
--Paul Roetzer, founder and CEO of SmarterX
This creates a vicious cycle: leaders who don't understand AI won't prioritize it, which means the organization never builds the infrastructure to adopt it at scale.
Consider the disconnect. Microsoft's CEO of AI, Mustafa Suleyman, recently predicted that in 18 months all white-collar work will be automated. Meanwhile, Roetzer sees a very different ground-level reality. He says he's been with companies who are so far behind that even if they've rolled out generative AI in its current form to all of their employees in 18 months, "I'll be shocked."
Human are the real barrier. The technology is not the bottleneck, people are.
The path forward starts with the basics. "For companies struggling to see ROI from AI, you have to go back to the basics, AI understanding and access." That means AI literacy at the leadership level, not just training programs for frontline employees. It means giving people actual access to the tools. And it means executives who are willing to accept calibrated risk.
None of this is glamorous. None of it involves frontier model capabilities or agentic workflows. But until organizations solve these foundational problems, the most advanced AI in the world won't matter because it will never reach the people who could use it.
Consulting firms as distribution. Will management consulting firms become the actual channel through which AI reaches large enterprises? OpenAI and Anthropic are betting on it. If the labs can't get in the door themselves, the consultants may become the most important players in enterprise AI adoption.
The leadership literacy tipping point. Whether enough C-suite leaders develop genuine AI fluency to break the approval logjam or whether the gap between AI-forward and AI-blocked organizations continues to widen.
Ethan Mollick on X — Original post on the enterprise AI adoption divide
Deloitte: State of AI in the Enterprise 2026 — Deloitte's mapping of the untapped edge of enterprise AI
Mollick: On Holding Back the Strange AI Tide — Ethan Mollick's deeper analysis of why organizations resist AI adoption
Heard on The Artificial Intelligence Show, Episode 201: This topic was explored in depth on the podcast with Paul Roetzer and Mike Kaput. Listen to the full episode for the complete discussion on enterprise AI adoption barriers, the Law of Uneven AI Distribution, and why humans remain the biggest obstacle to AI transformation. Listen →