Rippling launches AI Spend Console to curb runaway token costs

The company’s decision followed a March executive meeting where CFO Adam Swiecicki revealed that Rippling was burning 40% of its R&D budget on AI tokens. Analysis showed that a small group of employees drove 60% of that expenditure, with some engineers racking up $50,000 in monthly costs by defaulting to the most expensive frontier models for every task.

To regain control, Rippling built an internal AI gateway that routes prompts to the most cost-effective model for specific jobs. By switching to cheaper, high-performance alternatives like Z.ai’s GLM 5.2, the company maintained its volume of 600 billion tokens while cutting total costs by 63% compared to its April peak. Chief Product Officer Matt MacInnis emphasized that the new console tracks whether high spending actually correlates with improved output, such as lines of code or successful customer onboarding.

Beyond technical routing, Rippling appointed "AI captains" to guide staff and prevent inefficient usage. While the tool is now available to other companies, management remains cautious about universal access. The company maintains that if employees cannot prove that AI consumption drives measurable productivity, access to these tools may be restricted, signaling an end to the era of unrestricted enterprise AI experimentation.