AI has not increased our revenue.
But we're paying for it every month.
The subscriptions are active. Employees are using the tools. Every workflow connected to a model adds another source of token usage.
We call it an investment in productivity. The company pays for AI every month, even when its revenue hasn’t increased.
My expectation is that employees will eventually pay for some of that usage through lower salaries.
Your AI usage is part of what you cost
Your employer already pays more than your base salary. There may be a bonus, benefits, equipment, and other costs associated with employing you.
AI adds another expense: the tools you use and the tokens your work consumes.
A fixed subscription is one thing. Usage that grows with every prompt, agent run, and automated task is another. When AI becomes part of the daily workflow, that spending becomes part of the ongoing cost of getting the work done.
You may see the assistant as something that helps you do your job. Your employer also has to account for what it costs to provide it.
The question is whether the business is earning enough from that work to justify the additional expense.
A work subscription isn't automatically a revenue stream
A company selling an AI product to customers has a direct way to make money from it. Customers pay for access, and the company has to make the economics work.
For a company buying AI tools for internal use, the money initially flows in the other direction. It pays the providers.
That can be a useful expense. But the purchase itself doesn't create revenue. Neither does the number of prompts employees send or the amount of code they generate.
If the business is still selling the same products to the same customers for the same money, where is the additional revenue covering the AI bill?
That is why much of this spending feels to me like a subscription to a work tool rather than a revenue-generating investment. Companies shipping AI products have something to sell. Companies consuming AI internally still have to show what the spending returns.
Until they can, the cash keeps going out.
Productivity has to show up somewhere
You complete a task in an hour instead of three. That's useful. But your salary hasn't changed, and the company now has an AI bill as well.
The time saved needs to become something the business benefits from: more paying customers, a product delivered sooner, or an expense it can actually stop paying.
If the team simply produces more internal work while revenue stays flat, the employer is spending more without an obvious financial return.
At some point, someone will ask why.
Your compensation is where I expect the pressure to land
A company that commits to AI spending has to fund it. If additional revenue doesn't cover the cost, it will look elsewhere in the budget.
Employee compensation is one place I expect it to look.
That could mean lower salary offers for the next person hired into your role. It could mean smaller increases or bonuses. It could also mean pressure to accept an actual pay cut while the company continues paying for the tools it now expects you to use.
The message may be that AI makes the job easier or allows the same work to be done at a lower cost. Either way, employees could end up absorbing the expense through what they earn.
This is a prediction about where the pressure is heading. Buying AI doesn't automatically cause a salary cut, and a business that gets a clear financial return has a different calculation to make.
But when revenue hasn't increased and the AI bills keep growing, I wouldn't assume your compensation is protected from that calculation.
The company pays the provider. You may end up paying through your salary.
Comments