08/05/2026
I think a lot of businesses still imagine AI use at work looks something like this:
🤖 The company approves one AI tool
🤖 Employees log into that tool
🤖 Everyone uses it exactly as intended
But that’s not what’s happening at all 😅
New research has found that a huge amount of work-related AI activity is happening through personal accounts instead of company-approved systems.
And it works both ways.
Employees are also using company-paid AI tools for personal tasks.
People use whichever AI tool is easiest to access in that moment.
If someone already has ChatGPT open in a browser tab, they’ll probably use that. If they’re logged into a work AI platform already, they’ll use that instead.
Most of your staff aren’t sitting there thinking, “Is this my work AI account or my personal one?”
They’re trying to get something done quickly.
If the “official” AI tool is difficult to log into, slow to access, or awkward to use, people naturally drift towards easier alternatives.
The risk is what gets pasted into those tools 😬
Meeting notes, customer information, internal documents, ideas and plans…
Once that information ends up in a personal AI account, the business often has no visibility or control over it.
And if the employee leaves, that history leaves with them too.
I don’t think most people are doing this recklessly.
They’re doing it for convenience.
It’s why banning AI completely usually doesn’t work. People will still use it if