The latest MIT report (The GenAI Divide: State of AI in Business 2025), featured in Fortune, delivers a surprising message: 95% of major corporate AI projects fail to achieve profitability. Millions invested, endless hours of strategic planning and presentations… and in the end, very little impact on productivity or financial results. According to the report, only 5% succeed.
These figures contrast sharply with what happens in companies’ daily reality. In many cases, employees have already integrated AI into their routines, without waiting for the organization to authorize or regulate it. In practice, AI adoption isn’t written in strategists’ PowerPoints — it’s happening right on users’ desktops.
From a cybersecurity perspective, this is shadow IT in its purest form. It’s spontaneous innovation that creates value, but also opens the door to cybersecurity risks, potential leaks of sensitive information, and serious regulatory implications.
Sometimes, revolutions begin in silence. Most are born from the gap between what is and what could be. And although they’re often feared for the change they bring, what’s truly dangerous is remaining still when the world has already started to move.
The leadership of the future will depend on connecting these two worlds. And to do so, cybersecurity vision must be fully integrated. Only then will we move from hype to real transformation — with AI as a business ally, not as an invisible threat.
The true AI revolution has already begun — and it’s happening on our desktops.
This new season looks promising.