AI Now Powers More Than 95% of Code Development at Coinbase
According to Rob Witoff from Coinbase, the cryptocurrency exchange is increasingly relying on artificial intelligence for code execution while maintaining human oversight for strategic decisions and critical judgment calls.
More than 95% of the code developed at Coinbase now receives assistance from artificial intelligence, providing fresh perspective on the cryptocurrency platform's approach to AI integration after the company announced workforce reductions affecting 14% of its employees earlier in the year.
In May, Coinbase eliminated 700 positions from its workforce. CEO Brian Armstrong communicated to staff via email that artificial intelligence has "dramatically" transformed the velocity of operations, necessitating a shift to "return to the speed and focus of our startup founding, with AI at our core."
"Effectively, 100% of our employees are using AI on a daily basis here," Rob Witoff, Coinbase's head of platform, told Cointelegraph. "And close to 100% of our code, probably somewhere between 95% and 100%, is written by or with LLMs today."
This percentage represents a dramatic increase from Coinbase's earlier projections in February, when the organization reported that 40% of its codebase was AI-assisted, demonstrating the rapid acceleration of artificial intelligence integration throughout technology and cryptocurrency sectors.
However, Witoff emphasized that there exists a "wide spectrum" regarding implementation methods and dependency levels. He explained that while developing fundamental cryptographic systems predominantly depends on human expertise, prototype development has become entirely automated, with primary operational systems falling in between these extremes.
"For example, when we're writing core cryptography, we have industry-leading cryptographers that are meticulously researching and reviewing one line at a time."
"We're using AI quite a bit to test and make sure the code we've written is working the way it should, there's no vulnerabilities, we're verifying the math, but that's a much more manual part than where we're building internal prototypes, which is now effectively a 100% automated."
Smaller teams with "tastemakers"
This transformation has enabled Coinbase to restructure into more compact, experience-heavy teams, according to Witoff, with just two or three staff members now accomplishing tasks that previously demanded teams of 10 or more individuals.
"There were a lot of junior development roles that were impacted," he acknowledged regarding the May workforce reductions, although the layoffs affected various departments throughout the organization, encompassing marketing, legal, customer support and compliance functions.
"For those smaller teams to work, for people to have the taste, the judgment, I think a lot about people having the battle scars so they know how to point agents in the right direction."
According to Witoff, the typical Coinbase engineer currently operates between five to 10 AI agents simultaneously, with these artificial intelligence agents collectively completing coding tasks equivalent to approximately 1,200 human employees.
Looking ahead to 2030, Witoff projected that Coinbase could witness AI agents performing work equivalent to what would require 100,000 human employees, he said.
Coinbase joined a broader trend of technology companies reducing headcount this year as enterprise artificial intelligence capabilities expanded. Cryptocurrency exchange Crypto.com announced staff reductions of 12% in March, targeting positions "that do not adapt in our new world."
Block CEO Jack Dorsey revealed plans in February to reduce Block's workforce by 40%.
"We're already seeing that the intelligence tools we're creating and using, paired with smaller and flatter teams, are enabling a new way of working which fundamentally changes what it means to build and run a company," Dorsey said on X.
Additional cryptocurrency organizations that referenced AI-driven efficiency improvements when announcing workforce reductions this year included Kraken, Gemini, Messari and Dune.