Coinbase Says Over 40% of Its Code May Be AI-Generated, Could Reach 50% Next Month; Bitcoin Mentioned

• AI now writes 40%+ of Coinbase’s code

• Engineers use tools like Copilot, Claude Code and Cursor to refactor and ship code faster.

• Coinbase continues hiring ~350 roles, with 93 backend positions emphasizing AI skills.

Coinbase AI code: Over 40% of Coinbase’s code is AI-generated. See how this shifts engineering, hiring demand and developer productivity—read now for insights.

What percentage of Coinbase’s code is AI-generated?

Over 40% of Coinbase’s codebase is currently AI-generated, according to CEO Brian Armstrong. The company reports this proportion has more than doubled since April and may approach 50% soon, with human review and testing required before deployment.

How is Coinbase integrating AI into developer workflows?

Coinbase engineers routinely use AI coding tools—Copilot, Claude Code and Cursor—to assist with refactors, upgrades and new builds. These tools enable single engineers to complete work that previously required months in a matter of days, improving throughput while preserving code reviews and testing.

Change in the percentage of Coinbase’s code that is AI-generated. Source: Brian Armstrong

Brian Armstrong highlighted the rise in AI-generated lines of code and emphasized responsibility: not all areas can use AI-generated code and human oversight remains essential. Armstrong also indicated a workforce strategy focused on making employees “AI-Natives” rather than mass replacement.

Why is Coinbase still hiring engineers as AI adoption rises?

Despite increasing AI use, Coinbase lists roughly 350 job openings, nearly half targeting hard technical skills. About 93 backend engineering roles explicitly reference AI skills in job descriptions, reflecting demand for developers who can apply AI tools responsibly.

Is AI causing mass crypto layoffs?

Hiring in crypto has tightened since 2022, but large-scale layoffs have eased. Recruiters note AI is attracting talent and capital, but most industry experts—including PwC researchers and tech leaders—argue AI will boost productivity rather than trigger immediate mass job losses.

What safeguards does Coinbase apply to AI-generated code?

Coinbase requires review, testing and understanding of AI-written code before integration. The firm enforces human code review, targeted testing, and limits AI usage in sensitive systems to maintain security and compliance standards.

Frequently Asked Questions

How fast has Coinbase’s AI code share grown?

The AI-generated share of Coinbase’s code has more than doubled since April, rising to over 40% today. Management expects continued growth as teams expand AI usage where safe and efficient.

Will AI reduce job openings at Coinbase?

Coinbase still posts hundreds of openings; many roles now demand AI fluency. AI is shifting skill requirements rather than eliminating demand for engineers.

Key Takeaways

• Rapid AI adoption: Coinbase reports over 40% of its code is AI-generated, doubling since April.
• Human oversight: AI-generated code is reviewed and tested; not all systems accept AI-generated changes.
• Hiring shifts: Coinbase continues hiring, prioritizing backend and AI-capable developers; hiring patterns reflect AI-driven skill demand.

Conclusion

Coinbase’s move to leverage AI-generated code at scale signals a shift in how large crypto platforms build software. AI-generated code is accelerating engineering productivity while driving demand for engineers who can apply and supervise AI tools responsibly. Expect further hiring focused on AI skills and sustained human controls.

By COINOTAG • Published: 2025-09-04 • Updated: 2025-09-04

Relevant commentary and criticism referenced in this report include plain-text mentions of New York Post, David Sacks, and PwC researchers; all are presented as cited perspectives without external links.