Enterprise AI conversations have moved from experimentation to implementation. At Capgemini SPARK 2026, the focus shifted: from implementation to value realization. The gathering explored “Make It Real: From Action to Value,” and how organizations apply AI at scale for outcomes. ; SPARK revealed the same transition. Consequently, leaders ask how AI investments translate into outcomes. :

  1. Value Measurement Is Becoming The New Executive Imperative

SPARK’s clearest thread was from action to value. The fireside chat “The AI Value Imperative: Scaling Impact Across the Insurance Enterprise” featured leaders from Grange Insurance and Capgemini. The implications extend beyond insurance: as AI enters operations, success must connect to outcomes rather than activity. Models deployed, users onboarded, agents created, or tokens consumed describe activity – not enterprise value. Instead, the next phase connects AI investment with growth, productivity, customer experience, risk reduction, innovation, and better decisions. Consequently, enterprise AI is entering its value-management era

  1. The Industry Is Moving From Automation To Agency

Agentic AI, not surprisingly, was prominent. For example, AWS framed its session as “A Leader’s Guide to Agentic AI: From Automation to Agency,” while “The Agentic Economy Is Here” featured executives from Uber, Meta, and Universal Entertainment. Agentic systems can coordinate work toward outcomes. However, the economic question is whether greater machine autonomy produces better business outcomes.  The value puzzle still remains to be solved.

  1. AI Strategy is Not Just A Technology Strategy

AI transformation requires rethinking decisions, work, governance, and outcomes. The AI-first enterprise is becoming this generation’s digital-first enterprise, except AI increasingly participates as a resource, in the form of agents, tools and workflows, rather than merely digitizing workflows.

  1. Governance, Trust, Security, And Data Become The Foundation For Scale

As enterprises move from pilots to scale, governance, security, and trusted data become foundational. Moreover, controls are not solely compliance obligations; they enable confident deployment. The divide may emerge between organizations with sophisticated models and those able to operationalize them across trusted data, governance, security, and processes. Ultimately, AI advantage increasingly depends on the enterprise operating model surrounding the model.

  1. AI’s Economic And Workforce Impact Is Moving To The Foreground

Gita Gopinath, Harvard professor and former IMF chief economist, broadened the conversation to AI, productivity, inequality, the macroeconomic outlook, and labor markets. Consequently, AI’s implications extend to work, talent, skills, and productivity gains. For CEOs, therefore, it is increasingly a workforce, operating-model, and economic agenda.

  1. Power And Prediction: When The System Becomes The Product

Perhaps the most riveting session was the closing keynote from Ajay Agrawal, Professor of Economics of Machine Intelligence, focused on “Power and Prediction” and transitional AI. As AI makes prediction cheaper and more capable, the economic value of individual tasks will decline while value shifts toward the systems that orchestrate decisions, judgment, and action. The implication is not simply that enterprises must redesign work, but that leaders must rethink where value accrues as AI changes the economics of decisions. This raises the stakes for AI value management.

The Bottom Line: Stop Automating Yesterday’s Enterprise

Organizations spent years doing AI. Now, however, the next phase is becoming AI-first—connecting capabilities to outcomes while redesigning work and decision-making. Competitive advantage will come not from deploying more models or agents, but from reallocating capital, decisions, and talent faster as AI reshapes the economics of work. Leaders must look beyond productivity gains to determine how AI changes revenue, cost, risk, and experience, and how outcomes compound across connected processes. Therefore, enterprises must apply the Forrester AI Value Matrix to clarify the distinct ways AI delivers value across the business.

Share