Enterprise AI adoption isn’t going nearly as smoothly as most organizations planned. A recent McKinsey survey on AI shows that less than 40% of companies report enterprise-level financial impact from their AI investments, and this is despite years of strategic commitment. The most-cited culprits should sound familiar: complexity, workforce resistance, muddy ROI tracking, and overall organizational friction.
What is receiving far less attention is the organic adoption pattern that is quietly succeeding in the background. Even companies that aren’t pushing for AI adoption are beginning to lean into it anyway, and they’re not doing it because of any top-down mandate; they’re doing it with internal AI power users who are naturally introducing AI tools into some of the most important workflows in an organization: the workflows of executives.
The Strategic Position Executive Assistants Already Occupy
Workers who are embedded in the executive layer as a function of their jobs — such as executive assistants — have direct visibility into how decisions get made and know when priorities shift. They aren’t siloed into specific departments, and their positioning gives them insights that even the most well-planned AI rollout programs lack: immediate proximity to the decision-makers who can recognize the value.
EAs that use AI tools as part of their productivity skillset can produce cleaner meeting summaries in less time, create automated reporting pipelines, and never forget a scheduling change. The result is obvious improvements in day-to-day workflows for both the assistant and the manager or executive they are supporting.
Bypassing a pilot program, evaluation committee, or rollout plan, an executive assistant demonstrates the potential impact of their own curated AI tools without even trying. The senior leader notices that this support role is growing in effectiveness and better understands where others could benefit from similar integrations. Research from MIT Sloan on organizational tech adoption consistently points to trust and demonstrated value as the key drivers, with individuals who trust AI saying they’re over twice as likely to use it regularly. No mandate from above can create this kind of trust across an organization.
Why Top-Down AI Programs Stall
While AI is still a relatively new wrinkle in the enterprise AI stack, friction points in enterprise AI adoption are already well-documented. In a 2024 study by IBM, companies cited complexity and overall workforce readiness as the two biggest hurdles in scaling AI beyond simple pilot programs. Concerns over job displacement or elimination, as well as significant legal and compliance overhead, compound the difficulty for companies attempting to make these programs work.
The internal “AI champion” pathway sidesteps the majority of these issues, as there is no company-wide announcement that generates resistance or inherent apprehension to exploring the tools. It requires no workforce training program before workers get started, and the technology is adopted in a controlled environment with a single assistant, executive, and a set of repeatable tasks as the testbed. As enthusiasm over ease of use and capability of the tools ripples outward, the results speak for themselves.
How AI Fluency Develops in Practice
One all-too-common misconception about AI adoption is that integration with AI is a binary skill where someone either gets it or they don’t. In practice, AI fluency is a developmental journey that takes place with measurable progression, and that can benefit greatly from structured training in the same way that other professional capabilities do.
The earliest stage of fluency is tool familiarity, with individuals understanding the basic capabilities of AI models, developing consistent prompting approaches, and experimenting with tools like Notion for basic task automation that saves time. As proficiency grows, the scope of what can be achieved with these tools scales with it.
Multi-step workflow automation using tools such as Zapier is an exciting level-up for employees, allowing them to build systems that are reusable and customizable for specific job functions and preferences. More advanced applications of AI include integrated scheduling frameworks that account for preparation time and that can adapt when priorities shift.
At each stage, measurable changes in output can be observed, and those metrics — trackable as hours of time saved, reduction in turnaround time for certain tasks, and a lower volume of manual one-off tasks — easily make the case for broader adoption.
How Adoption Spreads Through the Executive Layer
The organizational tipping point that makes this model so reliable is when the executive (or executives) becomes an AI convert. Enterprise leaders exist within high-density peer groups within their own organizations, and those networks are a strong driver of technology adoption. A senior leader that observes a peer’s assistant or team doing more in less time by leveraging AI is the shortest path to that leader adopting the same tools and workflows for their own teams.
This turns the traditional push-based AI mandate into a pull-based adoption flow, where demand for this new way of doing things spreads organically, grounded in proof that is observable and quantifiable. It doesn’t require a company-wide memo or IT presentation, and resistance to adoption turns into excitement.
A Different Entry Point for AI Strategy
Organizations that want to be serious about AI adoption but are hitting a brick wall may find that embracing a new approach yields much stronger results. An internal AI power user — in the form of an AI-fluent executive assistant or other high-level support role — skirts many of the most problematic roadblocks of organizational mandates and can demonstrate to senior leaders how these powerful new tools can dramatically impact the way their employees work.
The companies that will make the most progress on AI over the next few years and beyond are unlikely to be those that launched what they believe are the most ambitious AI transformation programs. Instead, they’re more likely to be the ones that found low-friction pathways to clear results with a credibility layer that was obvious and impact that was repeatable and predictable. Assistants equipped with AI may be the most underestimated AI adoption vector in enterprise, but also one of the most powerful.
