By Lina Irawan 
The promise of artificial intelligence to revolutionize the workplace is widely accepted, yet many organizations are struggling to fully realize its transformative potential. A recent AI strategy session for a leadership team revealed a common paradox: despite investing heavily in AI trainings, providing access to multiple AI systems, and even deploying no-code platforms, a critical question went largely unanswered. When asked, "How many have built something with AI that changed how work gets done?" only one hand tentatively rose. Most leaders saw themselves and their teams as users of AI, interacting with it for one-off tasks, but not as builders – creators of personal agents, custom assistants, or reusable workflows. This stark distinction highlights a pervasive challenge: the "builder activation gap," which describes the significant chasm between the vast number of individuals who could build with AI and the surprisingly few who actually do.
The difference between merely using AI and actively building with it is profound. Assistance, while valuable, typically generates a one-time productivity gain. It helps draft an email, summarize a document, or brainstorm ideas for a specific, immediate need. This interaction is often ephemeral, disappearing after its purpose is served. Building, however, transforms that singular gain into a reusable tool, an automated workflow, or a tailored application that can scale across tasks, teams, and time. While many organizations diligently track how many employees log in to AI systems or engage in basic prompting, the more revealing metric—and indeed, the key to unlocking exponential value—is how many are actually constructing solutions.
Will Drover, Professor of Entrepreneurship & Innovation and Department Chair at the Neeley School of Business, Texas Christian University, and Founding Director of Neeley AI Forward, has observed this pattern consistently across executive education sessions and applied AI courses for working professionals. His findings underscore a remarkable but underutilized capability: nearly anyone who can articulate a problem or desired outcome in plain English now possesses the inherent ability to construct a functional AI assistant, application, or automation, often without writing a single line of code. Yet, despite this newfound accessibility, a significant majority remain on the sidelines, failing to translate their potential into practical, impactful creations.
Consider the transformative journey of Caroline Davis, Chief of Staff at Capital Factory, a prominent startup accelerator. Just a couple of years ago, Davis, like many others, viewed herself squarely as an AI user. When Drover posed his signature question about building a working tool with AI to her applied AI course, her hand remained firmly by her side. This self-perception reflected a broader industry trend where technology consumption far outpaced creation, especially among non-technical roles. The mental model was often that AI tools were sophisticated black boxes built by engineers for others to operate within.
Today, Davis’s professional landscape has been dramatically reshaped by the very technology she once only consumed. Many recurring facets of her demanding role are now efficiently managed through a suite of AI-powered tools she personally constructed. Her flagship creation is an intelligent agent she affectionately named "Sunny," after her daughter. Sunny is far more than a simple chatbot; it is a sophisticated operational partner, seamlessly integrating with Davis’s essential digital ecosystem, including her email, calendar, Airtable CRM, and Google Sheets. This integration allows Sunny to leverage approximately a dozen meticulously documented workflows that streamline critical processes, such as preparing executive briefs, meticulously tracking fundraising progress, efficiently onboarding new investors, and managing various administrative tasks.
The impact has been staggering. Data pulls and report generations that once consumed hours of Davis’s valuable time now take a mere 10 to 15 minutes. Furthermore, several of these automations are scheduled to run proactively, completing essential work even before Davis consciously requests it. A crucial aspect of Sunny’s design is the reusability and versioning of its workflows. Unlike transient, single-use AI interactions, Sunny’s components are designed to be improved, adapted, and reused, building a growing library of institutional knowledge rather than disappearing into the digital ether. Sunny even demonstrates a degree of coordination, regularly interacting with other specialized agents to achieve complex outcomes. Davis’s builder mindset has extended beyond her day job; she successfully applied the same AI-driven approach to rebuild a photography business she had operated a decade earlier, including the creation of its entire website, showcasing the versatility and scalability of her newfound skills.
Davis’s path from user to builder began with a manageable, foundational step. In Drover’s applied AI course, she leveraged natural language – the very medium of human communication – to construct her first functional AI assistant. This initial success was less about mastering complex technical jargon or coding syntax and more about cultivating a sense of agency and capability. Davis describes the ensuing shift in terms of confidence rather than technical mastery. That foundational experience, she explains, "built up my acumen as a whole and gave me the confidence to test out stronger AI models." This psychological pivot was critical: she stopped perceiving AI as merely a reactive conversational helper and started to envision it as a proactive, leverageable force capable of orchestrating recurring work. That initial, seemingly small build became the catalyst for the next, progressively leading to the comprehensive library of tools and workflows that underpin Sunny’s impressive capabilities. Along this journey, her self-perception underwent a profound transformation.
Despite her significant achievements, Davis maintains a refreshing humility, actively resisting the label of an "expert." "I don’t think I’m a power user by any means," she shared, "but most of my day is run through Claude at this point." Her experience with Claude, an advanced AI model, has become so ingrained in her daily operations that it underscores how accessible and impactful AI building can be, even for those who don’t identify as tech specialists.
Caroline Davis’s trajectory, while inspiring, remains an exception rather than the norm. Recent data paints a clear picture of this builder activation gap at a broader organizational level. According to Gallup, approximately half of all U.S. employees now utilize AI in their jobs at least occasionally, reflecting a significant baseline adoption. However, a deeper dive reveals that only 15% are daily users, suggesting that for many, AI remains a sporadic convenience rather than an integral part of their workflow. Further reinforcing this trend, a recent study published in HBR, based on an exhaustive analysis of 1.4 million AI interactions among over 2,500 KPMG employees, found that a mere 5% of the workforce qualified as "sophisticated users." These are the individuals engaged in iterative, higher-impact work that goes beyond casual prompting. For the overwhelming majority, AI usage is confined to providing assistance for one-off tasks, such as drafting basic emails or generating quick summaries of documents. While undoubtedly useful in the moment, these interactions are often disposable, lacking the cumulative, scalable impact of built solutions.
So, if the tools are accessible and the potential benefits so clear, where are all the builders? The answer lies in a complex interplay of structural barriers and a more insidious, often overlooked, psychological bottleneck: identity.
Structural Barriers:
- Governance and Security: Organizations, particularly larger enterprises, grapple with legitimate concerns around data privacy, intellectual property, and compliance. Restrictive IT policies, stringent access controls, and a fear of "shadow IT" can stifle experimentation and prevent employees from building tools that interact with sensitive data or core systems. The path to getting new tools approved can be cumbersome, discouraging initiative.
- Access to Tools and Data: While no-code platforms are becoming more prevalent, not all employees have equitable access to the most powerful AI models, specialized building environments, or the necessary datasets. Data silos and fragmented IT infrastructure can make it challenging for non-specialists to connect AI tools to the information they need to be effective.
- Time and Resources: The daily demands of a job often leave little room for experimentation and learning. Employees may perceive building AI tools as an additional, time-consuming burden rather than an investment in future efficiency. Without dedicated time, training, or clear mandates, building remains an extracurricular activity for a dedicated few.
- Incentives and Recognition: Traditional performance metrics rarely account for the creation of internal AI tools. Employees might not see a clear reward or career benefit for investing their time in building, especially if their efforts are not recognized, shared, or integrated into official processes. The focus remains on immediate output, not long-term systemic improvement.
The Hidden Bottleneck: Identity
Beyond these tangible obstacles, the most profound barrier is often psychological. Enterprise work has historically fostered a rigid division of labor: a select group of specialists (engineers, developers, IT professionals) are responsible for building and maintaining systems, while everyone else operates within the confines of those pre-built structures. This model made perfect sense when building required deep engineering expertise, complex coding skills, and specialized infrastructure knowledge. However, the advent of generative AI, large language models, and intuitive no-code platforms has fundamentally altered this equation. For a vast array of everyday work problems, building no longer demands specialized technical prowess.
What hasn’t evolved as rapidly as the technology itself is the prevailing identity within the workforce. Most employees, conditioned by decades of organizational structures, still perceive themselves primarily as consumers of technology, not as creators. They view "building" as inherently "not my lane," a domain reserved for those with specific technical titles. This identity is remarkably stubborn. As Herminia Ibarra’s seminal work on professional reinvention demonstrates, individuals rarely "think their way into a new self-image." Instead, they "act their way into it," with identity catching up only after sustained engagement in new behaviors. For many, the mental leap from "user" to "builder" is significant, requiring a fundamental shift in self-perception and perceived capability.
It’s crucial to acknowledge that not all AI building should be democratized. Complex systems, applications handling highly sensitive data, and customer-facing tools operating without human supervision still absolutely belong in the hands of specialized AI engineers, data scientists, and security experts. These require rigorous testing, robust security protocols, and deep technical understanding. However, a "huge share of everyday work problems" exists in a safer, more localized context. In these scenarios, the individual building the tool is often the same person who directly uses it, understands the nuances of the task, and is best positioned to assess its effectiveness and make iterative improvements. What often prevents these invaluable, localized innovations from flourishing is the ingrained belief that such creation is beyond their purview.
While AI has dramatically lowered the technical bar for building, making it accessible to a much broader audience, it has not yet succeeded in convincing most people that this capability is for them. Organizations are currently sitting on an untapped reservoir of potential builders, individuals who are precisely where Caroline Davis was just two years ago. The critical question for leaders today is whether they will recognize this latent capacity and actively cultivate it, or if they will allow this potential to remain dormant.
Activating a builder identity across the workforce requires deliberate, strategic intervention from leadership. Three core practices are essential for narrowing the builder activation gap:
- Foster a Culture of Experimentation and Psychological Safety: Leaders must actively encourage a mindset where tinkering, prototyping, and learning from "failures" are not just tolerated but celebrated. This means creating dedicated "sandbox" environments where employees can experiment with AI tools without fear of breaking critical systems or facing punitive measures for imperfect initial attempts. Regular showcases of internal AI innovations, even small ones, can inspire others and demonstrate what’s possible. Establishing internal communities of practice or "AI guilds" can provide a safe space for peer learning, knowledge sharing, and collaborative problem-solving, normalizing the act of building.
- Provide Targeted Training and Mentorship Beyond Basic Usage: Moving beyond generic "how to prompt" tutorials, organizations need to invest in practical, project-based training that teaches employees how to identify recurring problems, conceptualize AI solutions, and build functional prototypes using no-code or low-code AI platforms. This training should emphasize workflow design, data integration (where appropriate and secure), and the iterative refinement process. Pairing aspiring builders with more experienced internal champions or external mentors can provide invaluable guidance and accelerate their learning curve, helping them navigate technical challenges and overcome psychological barriers. Case studies of internal successes, like Caroline Davis’s Sunny, should be amplified to provide relatable examples and demonstrate tangible impact.
- Create Clear Pathways and Incentives for Building: For building to become an integral part of the organizational culture, it must be recognized, rewarded, and integrated into career development. Leaders should explore incorporating AI building skills into job descriptions for relevant roles, encouraging employees to dedicate a portion of their time to identifying and automating recurring tasks. Implementing internal recognition programs, "innovation challenges," or even "AI hackathons" can provide structured opportunities and incentives for employees to develop and showcase their creations. Crucially, ensuring that employees have authorized access to the necessary AI tools, data, and support within established governance frameworks is paramount. This includes providing clear guidelines on what can be built, with what data, and for what purpose, mitigating security risks while empowering innovation.
Organizations genuinely seeking to extract maximum value from their AI investments will find that simply tracking adoption numbers is insufficient. The true measure of success lies in their ability to narrow the builder activation gap, transforming a workforce of passive users into active creators. The journey requires a shift in mindset, a strategic investment in capability, and a conscious effort to dismantle both structural and psychological barriers. So, as Drover aptly suggests, leaders should revisit that opening question: ask your people what they’ve actually built with AI that has genuinely changed how work gets done, and observe the response. The critical work ahead is to ensure that the room no longer falls quiet, but instead buzzes with the energy of a workforce empowered to build.
The opinions expressed in Fortune.com commentary pieces are solely the views of their authors and do not necessarily reflect the opinions and beliefs of Fortune.
Will Drover is Professor of Entrepreneurship & Innovation and Department Chair at the Neeley School of Business, Texas Christian University, where he serves as Founding Director of Neeley AI Forward, a school-wide initiative, and as the Dean’s Advisor on AI. His work on AI adoption has appeared in MIT Sloan Management Review with coverage in the Wall Street Journal and Los Angeles Times. Drover teaches graduate courses on applied AI, runs executive education programs on AI strategy and leadership, and in practice holds ownership stakes in early-stage AI and robotics ventures.
