Token usage is fast becoming a central concern for enterprise AI buyers. As companies scale adoption of generative AI tools—particularly Anthropic's Claude—across their workforces, the cost of processing and generating text is emerging as a critical budget item and earnings-call talking point.
According to a review of earnings transcripts and analyst calls by WIRED, roughly 300 companies raised questions or concerns about AI token spending in April or May 2026. That represents a sharp increase from 93 companies that mentioned "token" in the same months a year prior—a threefold jump that underscores how central tokenomics has become to enterprise strategy.
**Divergent Approaches**
Companies are taking markedly different approaches to the challenge. 8x8, a Silicon Valley software maker that develops a communications platform for sales and customer service, has positioned itself as a cautious adopter. The company estimates it has saved approximately $5 million annually by canceling subscriptions to dozens of software and educational tools, relying instead on Claude for tasks like email drafting, customer feedback analysis, and code writing. Its annualized Claude bill remains "well below" that $5 million savings figure, according to Joel Neeb, the company's chief transformation and business operations officer. "It makes my chief financial officer happy," Neeb told WIRED.
However, 8x8 is beginning to wrestle with the trajectory. The company recently discussed the possibility of instituting usage caps with its CFO, prompted by growing internal adoption of Claude Opus 4.8—a newly released model that costs nearly 1.7 times more than an Anthropic offering from February. "Can we downgrade the model a little bit and still get the same outcome?" is now a central question, Neeb says. Access to Opus going forward might require employees to demonstrate that older models cannot accomplish the task.
In contrast, Baseball Lifestyle 101, a Long Island-based clothing brand expecting $250 million in sales this year, is taking an aggressive investment stance. The company told roughly 50 top managers to allocate the equivalent of about 20 percent of their salary to AI tokens monthly—a commitment expected to exceed $100,000 per month by year-end. Cofounder Bill Rom argues that the return justifies the spend. Claude recently identified that a retailer was running low on certain sizes of the company's ice-cream-patterned shorts, intelligence that led to a $1 million order. "That's a day and a half of work that can now happen in an hour or two that might make me eight figures of additional revenue over 12 months," Rom says.
**Industry Pressure**
The cost pressures are visible across the sector. Royal Bank of Canada's CEO disclosed that token usage surged 500 percent over the past six months. At Cisco, a third of employees use an internal AI chatbot daily, prompting CEO Chuck Robbins to describe token usage as "getting pretty, pretty crazy" on an earnings call. At analytics software developer Amplitude, some top engineers are spending thousands of dollars monthly on tokens, according to CEO Spenser Skates. Aaron Levine, CEO of Box, characterized token budgeting as "one of the most important" and "heated" topics facing the company.
**Managing the Volatility**
Executives cite several pressures unique to the AI cost problem. Pricing fluctuates unpredictably. New models—often more powerful and more expensive—are released monthly. And scaling adoption across entire organizations can create bottlenecks: productivity gains on one team can strain resources elsewhere. Some companies are developing or seeking systems to monitor token usage and automatically select the lowest-priced model for a given prompt.
8x8 has adopted a transparency-first approach, requiring all 1,800 full-time employees to check a dashboard tracking their and colleagues' Claude usage. "It's not punitive in the least; it's really just so that we all stay tightly packed in this journey," Neeb says. In May, product and customer success teams were the heaviest users; sales and finance teams among the lightest.
So far, 8x8 remains committed to AI expansion. Customer satisfaction and loyalty metrics are trending higher, and revenue has grown for four consecutive quarters. But the company is clearly preparing for the moment when token budgeting becomes as rigorous a discipline as any other line item in the P&L.