Okta COO: The Real AI Challenge is Managerial, Not Technical

Okta President and COO Eric Kelleher argues that the primary hurdle to AI adoption is not the technology itself, but a fundamental misalignment in corporate management. While many organizations are successfully experimenting with AI tools, they remain anchored to legacy management frameworks that prioritize human headcount and traditional organizational charts. Kelleher suggests that for AI to deliver meaningful value, managers must shift their focus from purely human-centric planning to a hybrid model that treats digital agents as legitimate, budgeted colleagues.

To bridge this gap, Kelleher advocates for a radical change in how companies approach budgeting. He proposes that firms move away from headcount-only planning and instead implement 'token budgets' for managers. By forcing leaders to account for both human labor and digital labor in their operational plans, companies can create a more accurate representation of how work is actually performed. This shift is intended to move the conversation beyond the fear of job displacement and toward a more productive focus on redesigning workflows to integrate AI agents effectively.

This managerial evolution is becoming increasingly urgent as the 'activation gap'—the disconnect between AI's theoretical potential and its actual productivity gains—continues to widen. Recent research from Cognizant indicates that while AI is disrupting jobs at a pace far faster than initially projected, the expected productivity surge has yet to materialize. With humans still required to oversee the vast majority of AI-assisted tasks, the challenge for modern leadership is not to replace the workforce, but to restructure roles to optimize the synergy between human expertise and machine efficiency.