Coinbase CEO Armstrong Gives Engineers Week to Master AI Coding, Targets 50% AI-Generated Code

Coinbase CEO Brian Armstrong fired engineers who failed to adopt AI coding tools within a week of his mandate, as the crypto exchange pushes toward generating 50% of its code through artificial intelligence by October.

“~40% of daily code written at Coinbase is AI-generated. I want to get it to >50% by October,” Armstrong posted on X, making the company among the most aggressive adopters of AI development tools in the tech industry.

Source: X/@brian_armstrong

The dramatic ultimatum came after Armstrong rejected internal estimates that AI adoption would take quarters to reach 50% among engineers.

“Originally, they were coming back and saying, ‘All right, over the next quarter, two quarters, we’re going to get to 50% adoption.’ I said, ‘You’re telling me— why can’t every engineer just onboard by the end of the week?’” Armstrong explained on John Collison’s “Cheeky Pint” podcast.

Several engineers were terminated after failing to provide adequate explanations for missing the deadline.

“I jumped on this call on Saturday and there were a couple people that had not done it. Some of them had a good reason, because they were just getting back from some trip or something, and some of them didn’t, and they got fired,” Armstrong revealed.

Armstrong acknowledged the “heavy-handed approach” faced internal criticism, but established clear expectations about the importance of AI at Coinbase.

“Some people really didn’t like it, by the way, that heavy-handed approach, but I think it did set some clarity at least that we need to lean into this and learn about it,” he said.

“Founder Mode” Mandate Triggers Immediate Workforce Transformation

Armstrong’s “rogue” Slack post bypassed traditional adoption timelines after hearing engineers might take months to embrace AI tools.

“I mandated it,” Armstrong told Collison. “I went rogue. I posted in the all-in Slack channel… I said, ‘AI’s important. We need you to all learn it and at least onboard,” he said.

Armstrong & Collison Source: YouTube

“You don’t have to use it every day yet until we do some training, but at least onboard by the end of the week. And if not, I’m hosting a meeting on Saturday with everybody who hasn’t done it, and I’d like to meet with you to understand why.’“

The Saturday accountability meeting included engineers with legitimate excuses, such as recent travel, alongside those without justification for non-compliance.

Coinbase’s AI strategy focuses on specific use cases rather than universal applications.

Frontend development, less-sensitive backend systems, and unit testing benefit most from AI assistance, while security-critical financial systems maintain human oversight requirements.

“Obviously, it needs to be reviewed and understood, and not all areas of the business can use AI-generated code. But we should be using it responsibly as much as we possibly can,” Armstrong noted on X.

“You don’t want people vibe-coding these systems moving money,” Armstrong cautioned. “You have to code review it and have the appropriate checks in place with humans in the loop.“

The company developed repository sensitivity matrices identifying criteria for AI tool usage, ensuring customer safety standards while enabling rapid development.

Notably, Coinbase’s monthly “AI speedruns” showcase successful implementations, with top-performing teams training colleagues on effective AI integration techniques.

“Every month we host, we call it an AI Speed Run where one of the engineers volunteers that month to run a training for how they’re using it,” Armstrong explained.

The productivity improvements enable single engineers to complete months-long projects in days.

“This has enabled profound success stories that weren’t possible 12 months ago, like single engineers refactoring, upgrading or building new codebases in days instead of months,” Armstrong noted.

Tech Industry Embraces Aggressive AI Adoption Despite Implementation Challenges

Armstrong joins other tech CEOs mandating AI adoption across engineering teams. GitHub surveys indicate 92% of programmers at major companies already use AI coding tools, with 70% reporting productivity advantages, validating Armstrong’s aggressive timeline.

However, implementation challenges persist across the industry. Stripe co-founder John Collison questioned the long-term sustainability during his podcast with Armstrong.

“It’s clear that it is very helpful to have AI helping you write code. It’s not clear how you run an AI-coded code base,” he said.

However, Armstrong acknowledged ongoing challenges. “I agree. I think we’re still figuring that out too.”

“You probably can go too far with it. You don’t want people vibe coding these systems moving money,” he emphasized.

The aggressive AI integration extends beyond engineering to other departments.

“We want to make sure it’s used not just in the engineering teams. It really should be any team. Design is using it heavily. Product managers. I think FP&A could even be using this,” Armstrong said.

Coinbase now includes AI as a participant in decision-making processes.

“We use a decision-making process called RAPIDS and everyone writes their input. We have a row now for AI that writes its input in as one of the people that help make decisions,” Armstrong revealed.

Notably, Armstrong also revealed that he makes extensive use of the tools. “Even as CEO, by the way, I use it a lot,” he noted.

Looking forward, while critics argue that the rapid adoption pace may compromise code quality and security, AI adoption continues to grow at a daily rate.

Just a few hours ago, former X engineer and startup founder Kache posted on X, “95% of my startup’s code is written by an LLM,” quoting a projection from Anthropic CEO Dario Amodei, that “by now, AI would be writing 90% of all code”.