Companies around the world are spending an estimated $1.5 trillion on artificial intelligence, and the results are already noticeable, with AI adding more than $400 billion to the U.S. economy alone. But beneath those headline numbers lies a little-known truth. Most (95%) of GenAI projects fail to achieve return on investment.
This disconnect is not a technology issue. It’s a conversion issue. And fixes don’t come from the boardroom or the IT department. It comes from the cubicle, the customer service desk, the HR team – the employees who know first-hand where the bottlenecks and opportunities exist.
Bottom-up AI movement
New data, based on our survey of 200 IT executives at billion-dollar US companies, reveals a quiet but historic shift in the way innovation happens. For the first time, we are driving the adoption of agent AI, a system that allows non-technical employees to take action, make decisions, and automate complex workflows on their own at a scale never seen before.
A whopping 91% of executives say they are playing a larger role in AI projects than non-technical staff have played in previous waves of technology adoption. These are not hypothetical use cases or innovation theater projects. The majority of these efforts (78%) are focused on solving real, enduring, everyday challenges. From automating repetitive workflows to surfacing insights buried in massive amounts of data across numerous systems, employees are using AI to reduce digital friction, return focus to passion projects, and drive business forward.
In our findings, 78% of leaders reported that agent AI is already driving significant change in at least one part of their operations. This is not about incremental change. It means rethinking the way we do work.
Changing corporate power structure
This change is not just technological. Organizational structures are changing. For decades, IT departments have been the gatekeepers of new technology, often serving as the “tallest tower” within a company. But data shows that that is changing rapidly. Our research shows that only 38% of executives currently believe that IT will be the department most responsible for AI innovation over the next three years.
The old concept of shadow IT, where teams bypass official channels and use their own tools, has long been considered dangerous or even reckless. But now the true nature of this approach is being recognized. This is a sign that employees across the company are hungry for solutions and willing to take the initiative to get them.
Other business teams such as operations, human resources, and customer service are also stepping up their roles as leaders of AI-driven change. This redistribution of power is making organizations more agile and responsive, and opening new avenues for career advancement. Four in 10 executives expect AI to create upward mobility for all employees, not just technical professionals.
The human side of AI transformation
This bottom-up change introduces new cultural complexities. 89% of employees embrace AI tools, but are keen to integrate them into existing workflows. Our research reveals that 65% support augmenting current processes with AI rather than forcing a complete overhaul.
This approach highlights important tensions between incremental improvement and bold change. The most advanced companies design AI around people, not the other way around. As one IT executive said in response to our survey:[Agentic AI is] This not only challenges the way we work today, but also opens new doors to smarter, faster and more collaborative ways of working. ”
Leaders need to recognize the cultural and structural implications of agentic AI, and successful companies will be those that embrace these changes while keeping people and purpose firmly at the center. Balancing immediate adoption with the potential for true innovation requires a delicate approach. Leaders must meet employees where they are and inspire them to envision a future where AI empowers employees and frees them to focus on monitoring systems and applying judgment in complex scenarios.
what happens next
First, leaders need to recognize that the most successful AI initiatives are not inherited from above, but come from the front lines. Organizations that allow employees to identify problems and experiment with solutions will outperform those that rely on mandates and one-size-fits-all platforms.
Second, the role of the IT department needs to evolve. Rather than acting as a gatekeeper, IT can be an enabler, providing guardrails, tools, and support while giving other departments the freedom to innovate.
Finally, leaders must address the cultural hurdles that come with major change. That means investing in education, building trust in new tools, and ensuring all employees, regardless of their technical background, have the opportunity to participate in the future of AI.
The true potential of AI lies not in the algorithms or hardware. It’s about unleashing the creativity, expertise and ambition of everyone in your organization. The future of enterprise AI is bottom-up, not top-down. And the companies that embrace this change will be the ones that truly transform.
Bhavin Shah is the CEO of Moveworks.
The final deadline for Fast Company’s World Changing Ideas Awards is Friday, December 12th at 11:59pm PT. Apply now.
