When Brian Armstrong, the Coinbase CEO, sat down for the “Cheeky Pint” podcast with Stripe’s John Collison on August 22, 2025, few expected the conversation to turn into headlines. But it did. Brian Armstrong openly shared that he had let go of Coinbase engineers who chose not to try out the AI tools the company had provided. Many headlines later summed it up as “Why Coinbase CEO fired engineers”.
Coinbase, the leading crypto exchange, had given its Coinbase engineers full access to GitHub Copilot and Cursor, with enterprise licenses covering everyone. Even so, some teams said AI adoption would take months. The Coinbase CEO did not accept that. He posted a message in the engineering Slack: learn the tools, get onboarded by the end of the week, or meet him on Saturday to explain. This moment is now often referred to as “Coinbase AI mandate explained.”
Meeting on a Saturday
That weekend, Brian Armstrong followed through. He sat down with those who had not onboarded yet. A few had good reasons such as being on vacation or traveling. But some did not. The ones without valid explanations were let go immediately, cementing the case of “Why Coinbase CEO fired engineers.”
Later, the Coinbase CEO admitted his approach was “heavy-handed.” He also acknowledged that it did not sit well with everyone. Still, his stance was clear. At Coinbase, AI adoption was no longer optional, forming the backbone of what many call the AI mandate.
Training, Teamwork, and a Bigger Goal
The number of dismissals was small. What mattered more was the message: adapt to AI adoption or risk being left behind. To strengthen this shift, Coinbase introduced monthly “AI speed run” sessions. Teams now share creative ways they are using AI tools, turning these sessions into moments of workplace innovation and collaboration.
At the same time, the Coinbase CEO reminded everyone that AI should not be treated as a replacement for human judgment. He agreed with Collison’s point that while AI can generate code, the real responsibility lies in how that codebase is managed.
Reports later highlighted Brian Armstrong AI strategy and his ambitious target: by the end of the quarter, he wants half of Coinbase’s code to be written with the help of AI. That goal goes further than what even Microsoft and Google are aiming for at the moment.
Bigger Picture: The Race at Coinbase and Beyond
Coinbase is far from alone in this shift. Companies like Duolingo, Microsoft, and Yahoo Japan already treat AI adoption as standard practice. Some even tie performance reviews to how well employees use AI, marking their own AI mandate.
The results, however, are mixed. Certain developers say AI slows them down. Others feel it limits their creativity, especially in high-pressure situations. Some even represent engineers resisting AI adoption. Yet the larger trend is undeniable. AI in crypto industry is becoming the default way of working in technology.
What This Really Means
Brian Armstrong’s AI strategy draws a line in the sand. AI adoption is not an experiment anymore. It is a strategic priority. His move reflects a wider cultural change happening across the tech industry, where using yesterday’s tools could cost tomorrow’s opportunities.
For Coinbase, the message is firm: adopt AI, learn with it, and grow with it, or risk falling behind. In other words, “Coinbase AI mandate explained.” For Coinbase engineers, it has become both a challenge and an opportunity for workplace innovation. While some are thriving, others represent the narrative of engineers resisting AI adoption.
In the bigger picture, AI in crypto industry is being shaped not just by big tech firms, but also by Brian Armstrong and his decisive leadership as the Coinbase CEO of the world’s most influential crypto exchange.
