Costa News Spain Breaking News | English News in Spain.
Ninety per cent of companies use AI software.
Credit: Iryna Imago/Shutterstock
Many companies like to boast about their billion-dollar AI programmes. According to a new study, the reality is far less impressive.
In reality, only 5% of the companies get real results. They are stuck in a loop of useless pilot projects.
In the meantime, a quiet change is taking place: The shadow AI economy.
Nine out of ten employees use AI “in secret”.
Project NANDA 2025, a project of MIT, found AI users in every corner. Over 90% of employees use popular AI tools such as ChatGPT, Claude and Copilot.
Their companies are unaware that they automate tasks, draft email, and analyse data.
Only 40 percent of companies subscribe to AI in an official capacity. The vast majority of AI is used in the business world unofficially, and therefore underground. While CEOs await the billion-dollar programme, employees quietly drive productivity.
Why shadow AI works
This shadow AI economy is thriving. Experts claim that it is all about how easy these tools are to use.
The employees get paid instantly, without any approvals or bureaucracy. They don’t have to wait weeks. It’s easy to customize, and the AI will adapt quickly to your work flow. All you have to do is click and ask.
The study says that the impact is real. Shadow AI handles 70% of email draughting, and 65% of basic analysis. This gives humans more time to do the complex work machines cannot handle (yet). Corporate AI pilots report almost no impact on the company’s operation.
Is Shadow AI the future for enterprise productivity?
In essence, the results are exactly opposite to what we anticipated. The employees are teaching the companies what works while the official programs fall short.
According to MIT’s research, companies must rethink the way they do business: “The organisations who build on this pattern will be the future of enterprise AI.”
Finally, there is another benefit. These free AI tools, which are widely used, don’t replace jobs in mass and do not change business models overnight as these costly projects did.
More tech news here.
