A growing number of workers faced with an ultimatum to learn and use artificial intelligence (AI) tools on the job are either quitting or being fired.
Last week, it happened at Coinbase Inc., where CEO Brian Armstrong said he terminated engineers who balked at using AI tools. The so-called “AI or the door” stance represents a new wrinkle in a potential escalation of hostilities between productivity-obsessed management and an increasingly nervous rank-and-file workforce eyeing an uncertain future in an era of AI agents and physical AI.
Fanning the paranoia are comments like those from Salesforce Inc. CEO Marc Benioff, who during a recent podcast call said he slashed the company’s customer-support staff to 5,000 from 9,000 “because I need less heads” when discussing the impact of AI on Salesforce operations.
“AI can be valuable, but when it’s pushed without proper planning, it slows work, creates mistakes, and risks costly damage,” Yijin Hardware CEO Gavin Yi said. “In manufacturing, a single wrong calculation can ruin expensive equipment. Leaders walk away when they’re told to use tools that don’t fit the job or face termination.”
While Yi sees potential in AI, he warns that forcing its use, especially with ultimatums like Coinbase’s, can undermine sectors where strict quality control is non-negotiable.
AI adoption often causes a short-term productivity drop, averaging 1.3% and sometimes higher, before improvements appear because of poor integration, retraining costs, and outdated systems, according to a U.S. Census Bureau study of tens of thousands of manufacturing firms.
Research also found that experienced developers took 19% longer to use AI tools, resulting in a momentary dip in productivity.
Corporate AI mandates aren’t radically different from previous technology waves, which required training among workers to get up to speed. But in forcing employees to embrace a fast-evolving technology at warp speed, some overeager employers are not just sacrificing short-term productivity but driving away experienced workers who either quit or — in Coinbase’s case — are fired.
“Being forced to use AI makes me want to leave the industry… I’ve embraced new technologies before, but this just slows me down and makes my work worse,” a frustrated senior software engineer with 15 years of experience said in a recent online post. “My company’s OKRs don’t justify the downsides, and now I just want to quit.”
Despite the obvious upside of AI use, several nagging issues keep cropping up in early use. Ironically, evidence suggests that forcing AI use may be undermining the very human capabilities that drive innovation and competitive advantage.
As AI becomes mandatory in corporate environments, forced adoption under threat of termination is multiplying errors rather than reducing them. Workers report that AI-generated mistakes are slowing projects, damaging equipment, and creating significant financial waste as employees struggle with systems they’re required to use regardless of effectiveness.
The problem intensifies with one-size-fits-all mandates that treat all departments identically. Companies implementing broad AI policies are discovering that industries requiring precision, safety protocols, and rigorous quality control don’t align well with blanket technology requirements.
Perhaps most damaging is the erosion of professional skills and job satisfaction. Experienced workers increasingly find themselves relegated to fixing AI outputs rather than applying their expertise to solve complex problems.
The focus on compliance over quality is undermining workplace excellence. When AI usage is a mandate, pride in craftsmanship suffers as workers shift from pursuing the best outcomes to meeting arbitrary adoption metrics.
“The Coinbase approach (use AI or get fired) represents everything wrong with forced adoption,” Yi said. “It prioritizes fear over thoughtful integration, and that’s exactly when you lose your best people.”
