By Kevin Thompson, CEO, Tricentis
Every technology wave comes with a familiar pattern. First comes excitement. Then rapid adoption. Then, usually much later than it should, an honest reckoning with what actually worked and what didn’t. AI is no different, except the gap between adoption and accountability is widening at an alarming rate.
On paper, enterprise AI often looks like a runaway success. In 2025, KPMG found that 85% of organizations have already started implementing AI in business operations. Further, a 35% average increase in productivity has been reported from those that have integrated AI agents into regular workforce operations. From the outside, it appears as though AI has crossed the chasm. Inside most organizations, however, the story is far more complicated. Despite massive investment and growing usage, the majority of AI initiatives are still failing to deliver meaningful, measurable business value. What’s often missing isn’t ambition or tooling, but quality intelligence, which offers the ability to understand where risk is introduced, how it propagates, and whether AI-driven systems can be trusted as they scale. Without early visibility into quality risk, organizations often fail to discover problems until they are already expensive, customer-facing, or have caused reputational damage. This tension plays out constantly in conversations with customers and peers. Teams feel pressure to move fast because everyone else is moving fast. AI promises major productivity in development, operations, and decision-making, and nobody wants to be the executive who slows things down. However, speed has a cost, especially when it outpaces our ability to understand where quality risk is being introduced and whether it is compounding across systems once AI is in production. Too many enterprises are building on top of AI-generated code, content, and logic they cannot fully explain, validate, or defend, while assuming that confidence and velocity are enough to carry them forward. It’s not enough to move fast. We must move fast with earnest discipline and governance leading at the helm
When activity masquerades as progress
One of the most dangerous aspects of AI adoption is how productive it feels. AI generates output instantly. Code compiles, dashboards populate, and answers appear. The surface signals all look positive. Despite this, anyone who has operated large, interconnected systems knows that plausibility is not the same as correctness. AI-generated work can pass basic checks and still fail in ways that impact the wider business. This failure can show up as a security gap, a compliance issue, or a business process that breaks under real-world conditions.
Software development is often where this problem becomes hardest to ignore. Generative coding tools are now embedded in everyday workflows, and developers are producing more code than ever. Yet many teams quietly admit that a significant portion of their time is spent catching mistakes that a seasoned developer wouldn’t make, correcting errors far down in the production cycle, or undoing AI-generated output before it breaks systems. The organization feels busier. The backlog moves faster. But outcomes do not always improve. In some cases, they get worse. That is not transformation; it is churn.
Why quality is the missing multiplier
The uncomfortable truth is that AI does not fail because it is too powerful. It fails because we ask it to operate without sufficient context, constraints, and accountability. Quality, in this sense, is not about perfection, but about intelligence and trust. In practice, this shows up as having early, actionable visibility into where risk is introduced and the ability to monitor it as systems evolve. Enterprises that treat quality as a downstream concern inevitably struggle to scale AI, because every shortcut compounds risk. When defects or quality gaps go undetected early, they don’t stay small. Instead, they multiply quickly when automation operates at enterprise speed. After countless conversations with customers, partners, and other industry professionals, what’s stood out to me is how consistently this lesson surfaces across fields. The organizations that see real returns from AI are not chasing novelty. They are building discipline into the way AI is adopted, measured, and governed. Quality engineering is not slowing them down; it is what allows them to move faster with confidence. By identifying quality risk early, continuously monitoring change, and understanding how systems behave end-to-end, they are turning AI from a liability into an advantage.
A leadership shift, not a tooling problem
For enterprise leaders, this moment demands a reset in how success is defined. AI initiatives should not be judged by how quickly they launch or how impressive they look in a demo. They should be judged by whether they hold up under scrutiny, scale without introducing fragility, and deliver outcomes the business can rely on. That requires governance that is practical, not theoretical, and accountability that is shared across technology and business leadership.
The organizations that get this right treat AI as part of their operating model, not a side experiment. They design trust into the system from the start, invest in the ability to detect risk early, and accept that some friction is the price of durability. Over time, that discipline becomes a competitive advantage, because it allows teams to innovate without constantly backtracking.
The solution: Pairing speed with trust
The billion-dollar AI problem is not that the technology is moving too fast. It is that too many organizations are moving without a clear plan for trust, including oftentimes struggling to secure confidence in the right AI partner or platform that can help advise them in the complex journey. AI will absolutely reshape how businesses operate, but this transformation will only be met with success for those willing to pair ambition with responsibility. When quality leads and accountability is clear, AI stops being a gamble and starts becoming what it was always meant to be: a reliable engine for long-term business value.
##
ABOUT THE AUTHOR
Kevin Thompson is Tricentis’ Chief Executive Officer and Executive Chairman of the Board. He joined Tricentis in April 2021.
Most recently Kevin served as President and Chief Executive Officer of SolarWinds until his departure from the company at the end of 2020. He previously served as Chief Financial Officer and Treasurer from July 2006 to March 2010, and as Chief Operating Officer from July 2007 to March 2010.
Prior to joining SolarWinds, Kevin was Chief Financial Officer of Surgient, Inc., a software company, from November 2005 until March 2006 and was Senior Vice President and Chief Financial Officer at SAS Institute, a privately-held business intelligence software company, from August 2004 until November 2005. From October 2000 until August 2004, Kevin served as Executive Vice President and Chief Financial Officer of Red Hat, Inc., a publicly-traded enterprise software company.
Kevin holds a B.B.A. from the University of Oklahoma.
