In a striking observation, Bill Briggs, the Chief Technology Officer of Deloitte, highlights a troubling disparity in corporate spending on artificial intelligence, with companies allocating 93% of their budgets to technology and only 7% to the people who will utilize it. This imbalance, he argues, neglects essential elements such as culture, workflow, and training, which are crucial for successfully integrating new technologies. Briggs points out that many organizations are falling into the trap of incrementalism, applying AI to existing workflows rather than reimagining them. This approach has led to a decline in trust among workers, with a recent Deloitte report indicating a 38% drop in confidence regarding generative AI between May and July 2025. Moreover, 43% of employees admit to bypassing company policies to use unapproved AI tools, further complicating the landscape.
Why do we care?
This is what happens when AI is treated like a software purchase instead of a management problem.
If 93% of the money goes to tech and 7% goes to people, you’ve already decided how this ends. You automate yesterday’s workflows, confuse your employees, and then act surprised when trust collapses and policies get ignored.
The shadow AI number is the giveaway. People aren’t being reckless — they’re being practical. They’re using tools that actually help them get work done, because the official program doesn’t.
For IT service providers, this is the warning shot. AI failures are about to show up as security incidents, compliance issues, and operational messes — and IT will be expected to clean it up.
The opportunity is to move beyond “AI enablement” and into adoption discipline. Workflow redesign. Governance that works. Clear decision boundaries. If you can help customers make AI useful and safe at the same time, that’s real value. If not, someone else will define the rules — and you’ll be stuck reacting.
MSPs that help clients make fewer bad decisions may be more valuable than those that simply help them adopt more technology.