Executives and technologists at the Wall Street Journal Leadership Institute’s Technology Council Summit on Tuesday pushed back against the fixation on short-term return on investment calculations for artificial intelligence, arguing that minute-by-minute productivity gains are a poor proxy for the technology’s strategic impact.

“For the most impactful business opportunity or products, you have to use AI to really double down on innovation versus productivity,” Sophia Velastegui, board director of BlackLine and former chief AI technology officer at Microsoft, told the audience. That view reflected a broader sentiment in the room: tallying isolated efficiency improvements — shaving seconds or minutes off routine tasks — produces comforting arithmetic but fails to capture AI’s capacity to create new revenue streams, reconfigure operations and reshape markets.

Participants described two distinct measurement approaches emerging in corporate practice. The first, bottom-up method aggregates discrete productivity wins across workflows — the familiar metric that equates saved human hours with cost reduction. The second, a top-down strategy, begins with business goals and asks whether AI contributes to achieving them, whether by enabling a new product, improving customer retention, accelerating time to market or transforming risk management. Advocates of the top-down approach urged technologists and line-of-business leaders to define success metrics before selecting tools, treating AI as one of several levers available to meet strategic objectives.

The debate highlighted practical obstacles to meaningful measurement. Attribution is difficult when AI is embedded in complex processes; benefits may be diffuse, delayed, or realized as qualitative shifts in customer experience. Executives noted that pilots often demonstrate local gains that do not scale, or that pass-through savings are absorbed elsewhere in the organization. Shadow deployments and inconsistent governance further muddy the picture, making it hard to compare projects or build a coherent portfolio.

Beyond internal accounting, speakers raised ethical and workforce implications. Deployments focused solely on automation can prioritize headcount reduction over reskilling and long-term competitiveness, they warned. Responsible measurement, participants said, should include nonfinancial indicators such as equity of outcomes, regulatory compliance and employee transitions, alongside revenue and efficiency metrics.

The summit dialogue underscored a growing consensus in corporate America: AI should be treated as a strategic capability rather than a cost-cutting plug-in. That shift demands new governance, clearer objective-setting and investment in change management. It also calls for patience. While executives can and should demand demonstrable returns, several panelists cautioned that meaningful value often accrues over years as models are refined, data ecosystems mature and organizations adapt.

For boards and executives, the takeaway was straightforward. Stop asking whether AI produced a few minutes of efficiency and start asking whether it moved the company toward its stated ambitions. When measurement begins with mission, leaders argued, AI investments are more likely to be judged by their ability to create distinctive advantage — and to justify the bets required to get there.