MIT recently published a study with a headline that stopped me in my tracks: 95 percent of enterprise AI deployments don’t affect the bottom line.
Think about that. For all the hype, budgets, and strategy decks, almost every attempt fails to move the P&L.
So the question is obvious: why?
The instinct is to say it’s the technology—that the models aren’t advanced enough, the data isn’t clean, the use cases aren’t clear. And sure, those things matter. But if you’ve lived through other big organizational rollouts—ERP systems, mergers, cultural change—you know the pattern. The spreadsheets always look good. The kickoff meetings are full of optimism. And then… the project stalls.
Which suggests the problem isn’t mainly technical. It’s human.
The Human Side of AI
AI doesn’t just alter workflows. It disrupts how people see their work.
If you’re an analyst, do you trust the tool or your own judgment? If you’re a manager, do you feel like you’re being asked to train the system that will one day replace your team? If you’re an executive, do you believe your people are ready for the disruption that’s coming?
These aren’t small questions. They cut to identity, trust, even fear. And they don’t show up on a Gantt chart.
This is where emotional intelligence—something we often talk about in leadership development but rarely in technology deployment—becomes essential.
EQ as Change Infrastructure
The EQ-i 2.0 framework breaks emotional intelligence into five composites. Each one has a direct line to whether an AI rollout thrives or fails.
- Self-Perception. Leaders who recognize their own anxieties about disruption are less likely to swing between hype and denial. Humility and curiosity matter here.
- Self-Expression. Clear, transparent communication—what AI will do, and what it won’t—builds trust. Spin does the opposite.
- Interpersonal. Empathy is critical. If employees are worried about being replaced, brushing off the concern won’t help. Listening will.
- Decision-Making. Reality testing keeps initiatives tied to business value, not vendor promises. Impulse control prevents scaling before proof.
- Stress Management. Change is stressful. Pilots stumble. Leaders who stay flexible and steady set the tone for resilience.
Notice what’s absent here: GPUs, data pipelines, coding sprints. The most significant predictors of success aren’t technical. They’re human.
Beyond AI
And this is the point worth sitting with. The MIT number isn’t just about AI. It’s about how organizations handle change, period.
Most big initiatives—AI rollouts, ERP deployments, cultural transformations—end up in what I call pilot purgatory: ambitious beginnings, expensive middles, quiet endings. And the reason repeats: leaders underestimate the emotional work of change.
What Leaders Can Do Differently
So what if we took EQ as seriously as we take technology?
- Assess. Utilize tools like EQ-i 2.0 to establish a baseline of where leaders stand in terms of readiness for disruption.
- Develop. If optimism is low, disengagement tends to follow. If empathy is weak, resistance festers. These aren’t soft issues—they’re rollout risks.
- Integrate. Bake EQ practices into the plan itself, not as an afterthought.
Because AI doesn’t fail in isolation, people often fail to integrate it effectively. And people succeed when leaders know how to guide them through the uncertainty.
Final Thought
The MIT study makes for a compelling headline: 95 percent of AI projects fail. However, the more interesting story lies beneath.
It’s not really about AI. It’s about us.
Technology doesn’t transform organizations. People do. And people only do it when they’re led with the intelligence that matters most in moments of disruption: emotional intelligence.