Father of Claude Code personally responds to the leak: the mistake was made by human manual steps, and the solution is to let AI handle more.

Coin World Network news: According to monitoring by 1M AI News, Boris Cherny, the father of Claude Code, responded on X to the incident of leaked source code, confirming that the cause was a human error in the deployment process: “Our deployment process has a few manual steps, and one of them wasn’t done right.” He said the team has made improvements and is adding more checks and safeguards. The context of this reply is rather telling. Last December, Cherny had posted that “in the past 30 days, my 100% contributions to Claude Code were completed by Claude Code.” After the leak occurred, some users dug up this post to imply that autonomous AI writing of code was the root cause of the incident. Cherny’s response directly denied this inference: what went wrong was precisely the human-operated manual steps. His reflection on the incident was: “Like other incidents, the counterintuitive solution is to find a way to speed things up, not to add more process. This time it’s more automation, and having Claude check the results.” A human error, and ultimately the solution it points to is more AI involvement.