Jacob Lauritzen, CTO of legal AI startup Legora, said tying AI tool token usage to employee rankings and performance evaluations can encourage "tokenmaxxing"—a behavior where employees artificially consume more tokens to appear more active on internal leaderboards, rather than improving actual work efficiency.
Prioritize output over consumption
He said on the podcast 20VC that this practice encourages employees to "burn tokens just to look better" without genuinely improving output. In contrast, a more effective approach is to have employees showcase how they used AI to complete projects and the specific efficiency gains achieved, through hackathons, internal demos, and similar formats.
Lauritzen believes companies should reward employees who are more efficient and produce higher output, rather than simply rewarding those who use AI the most. In his view, using AI itself is not the goal—the key is whether it leads to higher-quality work outcomes.
High-growth companies are still willing to pay for efficiency.
However, he also noted that for fast-growing companies like Legora, the opportunity cost of not using AI is equally high. If additional token expenditures can yield approximately a 20% improvement in efficiency, such investments remain practically meaningful.
Companies are beginning to tighten their AI budgets.
This statement comes as the tech industry’s approach to managing AI usage is shifting. Previously, some companies encouraged employees to experiment more with AI tools through leaderboards and internal dashboards, but as costs rise, an increasing number of organizations are concerned that such incentives may be backfiring.
- Uber has set a monthly spending limit of $1,500 for each AI tool.
- The Financial Times reported that Amazon has shut down its internal AI usage leaderboard.
- Cerebras CEO criticizes offering employees unlimited tokens
At a recent Bloomberg event, Andrew Feldman, CEO of Cerebras Systems, stated that not all tasks require invoking high-cost models; businesses should select more affordable open-source models based on task complexity to improve token efficiency.
From statements by Legora, Uber, Amazon, and Cerebras, tech companies are shifting their AI management focus from “encouraging maximum usage” to “pursuing tangible outcomes while controlling costs.”