According to TheRegister.com, more than two-thirds of corporate executives (68%) admit they’ve violated their own AI usage policies in the past three months. The survey from document company Nitro found that 97% of companies have invested over $1 million in AI to date, with 61% spending more than $10 million. Meanwhile, 70% plan to invest more than $10 million in the next 12 months, even as half of employees also admit to using unapproved AI tools. The research surveyed 103 C-suite executives and 900 managers across the US, UK, and Canada in October 2025, revealing that 82% of leaders believe their AI tools meet security requirements despite most breaking the rules.
Shadow AI starts at the top
Here’s the thing that really stands out: the people making the rules are the ones breaking them most frequently. Nitro CEO Cormac Whelan nailed it when he said “The shadow AI crisis starts at the top with C-suite executives who have built careers on finding workarounds.” Basically, when 68% of executives bypass the very tools they’ve invested millions in, they’re sending a clear message to everyone else: these approved systems aren’t getting the job done. And employees are following their lead – half of them admit to using unapproved AI tools too.
Massive investment, minimal governance
Companies are throwing insane amounts of money at AI while completely failing at the governance part. We’re talking about organizations where 97% have already spent over $1 million, yet most executives can’t even follow their own policies. There’s a huge disconnect happening here. Executives think their AI training is excellent (89% rated it that way), but only 63% of employees agree. And get this – 57% of employees feel low or no pressure to actually use AI tools. So we’ve got massive spending, minimal adoption pressure, and leadership that’s openly flouting the rules. What could possibly go wrong?
The hidden costs of AI
The financial reality might be even worse than the compliance mess. According to Gartner’s research, 73% of CIOs reported their organizations are breaking even or losing money on AI investments. For every AI tool companies buy, they should anticipate 10 hidden costs plus transition expenses. That’s a sobering thought when you consider that 70% of companies plan to invest over $10 million more in the next year. The Nitro survey data shows this isn’t just about software – it’s about fundamental business strategy failures.
What companies should do
Whelan and Microsoft are taking the same pragmatic approach: stop fighting shadow AI and start managing it. “Instead of fighting shadow AI adoption, IT leaders should focus on establishing guardrails rather than roadblocks and providing secure alternatives that executives will actually use,” he told The Register. The core issue seems to be that enterprise AI platforms are losing to consumer AI on speed, simplicity, and user experience. When even the C-suite would rather use unauthorized tools than the multi-million dollar systems they approved, maybe the problem isn’t the users – it’s the tools themselves. Adoption has to be earned, not mandated, and right now most enterprise AI isn’t earning it.
