The cryptocurrency exchange Coinbase has achieved an unprecedented level of automation in its software development process, utilizing artificial intelligence to generate over ninety-five percent of its entire software code. This was revealed by the platform's head, Rob Vitoff, in an exclusive comment to Cointelegraph. Remarkably, this figure has more than doubled in just a few months: as recently as February of this year, the company estimated the AI's share in development at around forty percent. According to Vitoff, AI-powered tools are now used daily by one hundred percent of the exchange's employees, fundamentally transforming traditional workflows. However, the degree of this automation directly depends on the complexity and criticality of the tasks at hand. For instance, the prototyping phase of new features is now fully automated, while core systems operate on a hybrid model where humans and machines complement each other. As for cryptography, manual work still prevails: experienced developers meticulously review every single line of code, while artificial intelligence serves merely as an assistant, helping to identify potential vulnerabilities and verify complex mathematical calculations.

Such a radical transformation of technological processes has directly impacted the company's human resources policy. In May of this year, Coinbase conducted a massive workforce reduction, laying off seven hundred employees, which accounted for fourteen percent of its total staff. The company's Chief Executive Officer, Brian Armstrong, publicly explained this difficult decision as a strict necessity to regain "startup speed and focus" by making a major bet on the ubiquitous implementation of artificial intelligence technologies. The integration of advanced algorithms has allowed the management to fundamentally restructure internal team dynamics. Today, small groups consisting of just two or three senior specialists successfully handle the volume of tasks that previously required the efforts of ten or more people. The layoffs primarily affected junior developers who lacked sufficient experience to effectively supervise AI outputs, as well as a significant portion of employees in the marketing, technical support, and compliance departments, whose routine functions have been successfully delegated to automated systems.

The sheer scale of this digital transformation is clearly demonstrated by the company's new internal efficiency metrics. Currently, every Coinbase engineer is supported by five to ten active AI agents that perform auxiliary and routine operations around the clock. Collectively, these virtual assistants execute a volume of work that is entirely equivalent to the labor of one thousand two hundred human employees. The company's leadership has set even more ambitious goals for the future, projecting that by the year 2030, this figure will grow to one hundred thousand virtual workers, which will allow for a complete overhaul of their business model. It is worth noting that Coinbase is far from the only technology company heading in this direction. Recall that in April, the team behind the popular messaging app Snap announced the replacement of one thousand employees with neural networks to save five hundred million dollars, and in May, the analytics platform Dune announced a deep restructuring and a twenty-five percent reduction in its workforce, confirming the global trend of substituting human labor with advanced algorithms.