Technology doesn’t meet resistance. People do.
MIT Sloan’s political lens frames the organization not as a flowchart but as a web of alliances, ambitions and competing interests. The political lens asks who holds influence, whose authority is threatened and whose cooperation is essential. When AI adoption enters that web, it doesn’t just change workflows — it reshuffles power. Roles gain relevance or lose it. Decisions that once required a team can now be made by one person with the right tool. And the people whose standing depended on controlling information, approvals or institutional knowledge feel it immediately.
Don’t Lead With Mandates — Build Coalitions From the Ground Up
Top-down directives rarely win political battles. Thomson Reuters understood this when it co-led its internal AI adoption strategy across HR and technology rather than handing it to either function alone. Rather than issuing a company-wide mandate, it identified 400 “AI champions” — early adopters and enthusiastic users spread across all functions and seniority levels — and formalized their influence. Champions were expected to share use cases, model behavior and advocate peer to peer daily. By November 2024, employees had completed the company’s AI Foundations course and were actively using generative AI tools. The political insight: influence travels faster through trusted peers than through policy memos.
Middle Management Is the Battleground
The middle layer of most organizations is where AI adoption most reliably stalls. McKinsey’s research is direct on this point — managers and senior practitioners who set the cultural tone are often the most resistant to change because their current methods work reasonably well and the learning curve feels daunting. One CEO of a major conglomerate addressed this by asking 100 business leaders to each personally sponsor an AI project with specific revenue or cost targets — targets that had to be reflected in the following year’s budget. Ownership became non-negotiable. When managers have skin in the outcome, resistance shifts to investment.
Leaders Must Use AI — Not Just Champion It
The most damaging political signal an executive can send is delegating AI adoption to IT while changing nothing about their own work. A McKinsey survey found that when respected leaders publicly share their AI learning journeys — including what they don’t yet know — it lowers the psychological barriers for everyone else. When a Chief Marketing Officer uses AI-driven analytics in her own decisions or a sales manager runs AI forecasting in weekly reviews, it sends a signal that no policy document can replicate: this matters and it starts here.
Power Follows Behavior
AI doesn’t fail because the tools are inadequate. It fails because the political conditions for adoption were never built. The organizations advancing furthest are those where leadership mapped the alliances before launching the initiative — identifying who needed to be won over, who needed a seat at the table and who would quietly block progress if left out. That work belongs squarely in Lens 2. And until it’s done, AI remains exactly what opponents want it to be: optional.
