Paul Smith, Anthropic’s chief commercial officer, said the leak was “absolutely not breaches or hacks,” and the mistakes have been addressed. “They’re part of the incredibly rapid release cycle that we’ve had around Claude Code,” Smith said in an interview Wednesday.
Anthropic takes the issue “incredibly seriously,” he added, noting that the firm has “all the right people focused on addressing it.”
The unintended release marked Anthropic’s second security slip-up in a matter of days, compromising approximately 1,900 files and 512,000 lines of code related to Claude Code. Last week, Fortune separately reported that Anthropic had been storing thousands of internal files on a publicly accessible system, including a draft blog post that detailed an upcoming model known internally as both “Mythos” and “Capybara.”
In a series of posts overnight on He said the company has already “made a few improvements to the automation for next time,” with plans for “a couple more on the way.”
The code included at least eight unreleased features, according to Abhishek Tiwari, an entrepreneur who quickly coded a website to track the leak using AI. Among them, he listed:
Kairosa setting that allows Claude to run in the background.
Coordinator Modewhich allows AI to break a task into pieces and delegate them to individual workers.
Auto-Dreamin which Claude reviews what it learns and organizes notes into clean, structured memory files.
ultraplanwhich allows the creation of a separate cloud instance that explores and plans up to 30 minutes at a time.
Anthropic did not comment on the features or its plans. “We’re always experimenting with new ideas. 90% don’t ship because we don’t think they’re good enough experiences,” Cherny said in a response to a post on X.
Also Read: Trump administration gears up for new tariffs on select drugmakers
