How CEOs Can Unlock Gen AI Value: A Five‑Step Change Management Playbook

Generative AI (Gen AI) is rapidly entering the workplace, yet many leaders struggle to translate pilots into tangible value. McKinsey’s latest report, Re‑configuring Work: Change Management in the AI Era , provides a five‑step roadmap for CEOs to guide organizations through this transition.

Step 1: Define an outcome‑based “North Star” goal, not just a tool

Gen AI should be seen as a capability rather than a mandatory tool. CEOs need a simple, inspiring North Star that can evolve with technology, describing how AI agents will create value, give a competitive edge, and influence the talent lifecycle.

Successful transformation requires a well‑funded change plan and end‑to‑end workflow redesign, moving from single‑task AI agents to coordinated “agent clusters” that deliver full business outcomes. Parts of the organization can become a “Minimum Viable Organization” (MVO) where AI agents handle most work and humans only verify outputs, while other areas retain augmented teams for high‑touch functions.

Step 2: Build trust through data access, governance, and enterprise intelligence

Scaling Gen AI depends on trust. Without confidence in AI outputs, employees will reject decisions. Companies that invest in AI‑driven trust activities see roughly double the probability of >10% revenue growth.

Data access must become a core change‑management workflow. Gen AI can consume unstructured data and uncover hidden insights, but training models on both structured and unstructured data remains challenging. New monitoring frameworks are needed to guard against hallucinations and bias.

Governance actions include establishing an AI oversight committee, acceptable‑use policies, compliance guidelines, and “human‑in‑the‑loop” checkpoints. For example, Morgan Stanley trained a Gen AI assistant on 100,000 research reports, achieving a 98% adoption rate after rigorous quality checks.

Step 3: Reshape workflows toward AI‑centric teams

Gen AI must be embedded in the core of workflows, not added as an afterthought. A “two‑person‑one‑team” model pairs business and technical staff. Evolution occurs in three phases: (1) independent agents handle discrete tasks; (2) agent clusters manage end‑to‑end processes with human supervision; (3) agent clusters become fully autonomous MVOs while humans focus on high‑value work.

Employee participation, formal training, and daily integration are critical. McKinsey’s Lilli platform shows 92% employee usage, with 74% regular use, saving 30% of time on information‑seeking tasks.

Step 4: Rethink organization structure, blend MVO and augmented teams

Leaders must decide which functions become streamlined, highly automated MVOs (e.g., invoice processing) and which remain augmented teams that combine AI analysis with human judgment (e.g., sales or customer service). This hybrid structure maximizes productivity while preserving the human touch.

The report notes that such a mixed model can unleash employee potential and control costs.

Step 5: Empower employees to become change agents

Successful transformation requires active employee involvement. Companies that engage more than 7% of staff double their success rate; top performers involve 21‑30%. Identifying Gen AI “super‑users,” especially millennial managers (35‑44 years) with high expertise, helps drive adoption and mentorship.

Examples include McKinsey’s adoption team and Lilli Club, and Singapore Telecom’s AI Acceleration Academy, which partnered with universities to train 10,000 employees.

Conclusion

The McKinsey report gives CEOs a practical roadmap for the Gen AI era: by following the five steps, organizations can move from pilot projects to sustained value creation, enabling employees to collaborate with AI, eliminate repetitive work, and focus on innovation, while managing cultural resistance.