Companies track AI use as a management signal

Some executives are starting to treat AI use less like an information-technology expense and more like a management tell. At Box, Chief Executive Aaron Levie said he looks to a Slack channel to see who is actually using AI tools, instead of rewarding the people who spend the most tokens. (businessinsider.com) Levie told Business Insider that Box tracks token usage but does not run a leaderboard, saying that would create “hilarious outcomes.” He said a dedicated Slack channel gives him a better read on which employees are experimenting with AI agents in real work. (businessinsider.com) That view lands in the middle of a Silicon Valley argument over “tokenmaxxing,” the practice of pushing employees to consume large volumes of artificial-intelligence tokens, the units that providers bill for when models read and generate text. Levie said he is fine with some wasted tokens if it means workers are trying new things, while Nvidia Chief Executive Jensen Huang has separately argued he would be alarmed if a highly paid engineer was not using large amounts of tokens. (businessinsider.com) Hiring teams are moving in the same direction. LinkedIn research published in January 2026 found that 93% of recruiters planned to increase their use of AI this year, and 66% said they expected to use it more for pre-screening interviews. (hrdive.com) The same LinkedIn findings, cited by CNBC and HR Brew, said 59% of employers reported that AI had helped them surface candidates they might not have found otherwise, and 70% of recruiters said AI could help them have more valuable pre-screening conversations. Those are the kinds of workflow gains companies are now trying to measure, whether through software dashboards or behavior inside internal chat tools. (cnbc.com) The shift comes as recruiters say the job has become harder, not easier. HR Brew reported in January that recruiters were planning heavier AI use while candidate matching remained difficult, and in April it reported that firms were turning to AI pre-screening tools to handle overloaded applicant pools. (hr-brew.com) Companies adopting these tools are also running into rules that did not matter much a few years ago. New York City’s automated employment decision tools law bars employers and agencies from using covered hiring tools unless they complete a bias audit within one year, publish information about that audit, and provide notices to candidates or employees. (nyc.gov) Federal regulators have also warned that old discrimination law still applies when software makes or assists employment decisions. The Equal Employment Opportunity Commission says workers remain protected if AI systems discriminate on the basis of race, sex, disability, age, national origin, religion, color, or genetic information. (eeoc.gov) Illinois added another layer on January 1, 2026, with new notice obligations tied to AI use in employment decisions under amendments to the Illinois Human Rights Act, according to summaries of the state’s draft rules. As companies watch who is using AI and where they use it, the management signal is starting to look like a compliance signal too. (hinshawlaw.com)