Beware “trophy-style” AI adoption that flatters usage

trophy-style AI – As enterprise leaders pour money into generative AI, a new benchmark survey finds most still struggle to convert investment into real value—largely because of human obstacles like culture and change management. Executives are now rolling out workforce initiati

For many enterprises, the problem with generative AI is not the technology—it’s the scoreboard. Companies can see activity ramp up, seats get filled, logins increase, and deliverables appear. Yet the value leaders were promised still hasn’t arrived in the way they envisioned.

In this year’s AI & Data Leadership Executive Benchmark Survey. 93% of executives leading AI and data efforts identified human issues around culture and change management as the primary obstacle to adoption. Bob Sternfels. McKinsey Global Managing Partner. made the point even more bluntly on HBR’s IdeaCast: “Half if not more of the secret sauce” in getting value from AI. he said. “is organizational change. as opposed to technology implementation.”.

That message has pushed many leaders to launch initiatives over the past several months to drive AI adoption across their workforces. Some are rolling out hackathon programs and prizes for innovative uses. Others track weekly logins and token consumption, treating those signals as a stand-in for performance.

The intention is understandable. People side matters. But the way adoption is measured can quietly decide whether those initiatives build capability—or build theater.

In a warning that lands in exactly the space between effort and evidence. Maril MacDonald. founder and CEO of Gagen MacDonald. describes a “trophy-style” approach: an AI adoption strategy that is performative and centered more on usage than results. She calls it “a performative approach focused more on usage than results. ” with leaders awarding “participation trophies over proof of impact.”.

MacDonald argues that employees prioritize what leaders model, incentivize, and reward—and that shallow metrics can do real harm. It’s especially tempting in moments when internal adoption metrics look underwhelming. When deliverables are celebrated simply because they were made with AI. or when employees are rewarded for integrating it into workflows. the organization can start treating any increase in AI usage as a win.

At her firm, the concern is that activity can masquerade as impact. MacDonald frames the danger as an illusion of progress—“a dangerous illusion of progress. where activity masquerades as impact.” The reward structure can end up favoring output over outcomes. leaving employees “less equipped to meet business needs than prior to AI.”.

There’s also a practical reason not to confuse the two, she says: not all AI use cases are equal. In her view. the work that comes out of AI tools can vary dramatically. and “workslop” and research on cognitive atrophy point to how certain patterns of use can fail to translate into better performance.

The sequencing of these facts is hard to ignore: executives say the biggest obstacle is culture and change management; leaders respond by encouraging adoption through programs and usage proxies; and then MacDonald warns that rewarding the wrong signals can leave organizations with a culture of shallow adoption rather than measurable ROI.

MacDonald’s counterpoint is that impactful adoption does not look identical across companies or roles. For some, it means deepening the quality of the same work product. For others, it means increasing output without sacrificing quality. For still others. it means getting the same work done in less time and repurposing time and energy toward new questions and tasks.

She also lays out a common design principle for initiatives: reverse-engineer them from the larger business strategy. The best programs, she says, will be built around metrics that connect to that strategy.

That process, she argues, should begin with clarity and specificity about big-picture questions. What does value look like for the organization?. How can different roles change to better deliver it?. And as adoption initiatives get designed. leaders can’t lose sight of these framing questions because they determine what gets modeled and encouraged.

The goal, in her telling, is business impacts that come from usage—without turning speed or deep integration into the entire point. When strong use cases are showcased, the emphasis should remain on meaningful organizational outcomes driven by the use case, not just the visible adoption of the tool.

Her warning closes on urgency: given how much companies have already spent. and plan to keep spending. on generative AI. the mismatch between usage and impact is not a mistake leaders can afford. “Because while the employees who create the most impact with AI will certainly use it frequently. it’s a mistake to think of usage as synonymous with impact.”.

In an AI race where dashboards can light up quickly, the real test is slower: whether leaders can build a culture where AI use translates into outcomes the business can actually feel.

generative AI adoption AI & Data Leadership Executive Benchmark Survey culture change management organizational change workforce AI initiatives hackathon programs token consumption metrics AI usage vs impact ROI cognitive atrophy