Corporate America is starting to ration AI, and it’s affecting marketing teams. Axios and The Wall Street Journal report that some enterprises have burned through their entire annual AI budget in just a few months. Others have watched AI spending double or triple with little warning.
For marketers, if your team is running AI tools for content, campaigns, reporting, or outreach, the cost dynamics have shifted dramatically, and the budgets most organizations set in late 2025 were not built for AI agents.
What’s Actually Driving the Spike
Unlike a chatbot that answers a single question, AI agents work through tasks in multiple steps, and every step consumes tokens. Tokens are the units AI providers use to measure and charge for compute. An agent drafting a content brief, researching prospects, or pulling campaign performance data isn’t making one request. It’s making dozens, sometimes hundreds.
As Goldman Sachs noted in a May report, “agentic AI requires a lot of tokens because many queries are repeated in sequence. It’s like taking a simple chatbot request and blowing it up 10-fold, 20-fold, 50-fold.” Token consumption is projected to multiply 24 times between 2026 and 2030. For marketing teams increasingly relying on agentic workflows, that growth curve will show up in their budgets well before 2030.
Marketing Teams Need AI
Marketing departments use a lot of AI power in their work: Content creation, personalization at scale, campaign analysis, social copy, email sequences, SEO research. All of these AI-assisted workflows have expanded rapidly, often without a corresponding update to budget or governance.
The problem is compounded by visibility. Most marketers don’t know how many tokens they’re consuming or what’s driving the bill. Usage varies wildly across team members, and the tools themselves don’t make it easy to connect spend to outcomes. You might know that your team hit its monthly limit. But you’re not sure which work produced the most value per token.
How Should Marketers Respond to Rising Costs?
The answer isn’t to cut AI access because that trades one problem for a bigger one. Instead, marketing leaders need to bring the same strategic thinking to AI spend that they apply to any other tech investment.
A few places to start:
- Audit before you restrict. Before setting new limits, work to understand where tokens are actually going. Which workflows seem to be consuming the most? Which are producing measurable marketing outcomes? That picture should drive decisions, not raw usage numbers.
- Match the model to the task. The most capable, and in turn most expensive, model isn’t always the right one to use. A frontier reasoning model is overkill for generating a social media post or blog. Build a culture of deliberate model selection to help reduce costs without reducing output quality.
- Where is AI essential? Token usage is the wrong KPI. What matters is what the AI-assisted work produced: time saved, content produced, pipeline influenced. Teams that connect AI activity to valuable marketing outcomes are better positioned to justify spend and make smarter investment decisions going forward.
The marketing leaders who navigate budgets well will have a solid strategy behind it.
Watch: Paul and Mike discuss growing AI Costs and Marketing Budgets
On Episode 217 of The Artificial Intelligence Show, co-hosts Paul Roetzer and Mike Kaput discussed why AI costs are rising, what’s driving the strain on enterprise budgets, and how marketing leaders should approach decisions on AI usage in the future.
