Block has handed over 15% of its code to AI: Builderbot rewrites the rules of development
Blockchain giant Block, led by Jack Dorsey, has taken a decisive step toward automating development. The company has deployed its own AI tool, Builderbot, which already generates about 15% of the firm's entire software code. This is not just a supporting assistant — it is a full-fledged orchestrator capable of taking on a significant portion of routine work.
How Builderbot Works
The tool is integrated into Slack: a developer simply tags the bot and describes the task. Builderbot independently finds errors, suggests fixes, or creates new features. The key difference from standard assistants is access to Block's entire codebase. An engineer from the Cash App team can make changes to Square services they have never worked with before. The bot automatically pulls tasks from Jira, creates branches in the repository, writes code, and submits Pull Requests.
Numbers and Scale
The scale is impressive: Builderbot performs over 200,000 operations per day and closes about 1,500 code merge requests weekly. "What used to take months now takes days. The bot handles the routine and environment setup, allowing engineers to focus on solving complex problems," noted Brad Axen, head of Block's AI capabilities department.
Security and Open Source
Importantly, the tool only works with source code and system configurations, without access to customer data or payment information. Builderbot is built on the goose framework, which Block contributed to the Agentic AI Foundation. The development also used the Model Context Protocol (MCP), created in collaboration with Anthropic.
The company is confident that the shift from writing code with AI to "native" engineering processes based on neural networks will become the main trend in the IT industry. And this is not just a forecast — it is already a reality where 15% of code is written by machines.
As an analyst, I see in this not just optimization, but a paradigm shift. If AI was previously a tool for acceleration, it is now becoming a full-fledged team member. Block is paving the way that many tech giants will likely soon follow. The question is not whether neural networks will replace developers, but who will adapt to the new conditions faster.