Trending Now - Anthropic Admits Source Code Leak, Calls it Human Error

WHO:

Inc - Anthropic's CEO is Dario Amodei.

CNBC - Anthropic, a company specializing in artificial intelligence, confirmed the leak of its source code.

WHAT:

Inc - The leak provided competitors with a roadmap to one of the fastest-growing products in AI, revealing the inner workings of Claude Code.

Business Insider - An Anthropic spokesperson reiterated, "This was a release packaging issue caused by human error, not a security breach."

Engadget - The leaked code was uploaded to a public GitHub repository and copied more than 50,000 times before Anthropic took action to address the leak.

Engadget - Evidence of a feature called "Proactive mode" was found, which will allow Claude Code to work even when the user hasn't prompted it.

Engadget - A potential crypto-based payment system was discovered that could allow AI agents to make autonomous payments.

Engadget - A Tamagotchi-like virtual companion that "reacts to your coding" was also mentioned in the leaked code as a possible feature.

Engadget - The leaked codebase was uploaded to a public GitHub repository, where it was subsequently copied more than 50,000 times.

Engadget - The leak exposed more than 512,000 lines of code and 2,000 TypeScript files.

Engadget - Evidence of a feature called "Proactive mode" was found, which will allow Claude Code to work even when the user hasn't prompted it.

Engadget - A potential crypto-based payment system was discovered that could allow AI agents to make autonomous payments.

Engadget - A Tamagotchi-like virtual companion that "reacts to your coding" was also mentioned in the leaked code as a possible feature.

Inc - The leak revealed the "Kairos" updates for Claude Code, which include features allowing Claude to work in the background and a "dream mode" for memory consolidation.

Inc - Following the leak, Anthropic's Claude chatbot experienced a surge in downloads, briefly reaching No. 1 in the US Apple App Store.

Inc - The leak of Claude Code's source code provides competitors with insights into the product's development strategy.

Business Insider - An Anthropic spokesperson stated, "Earlier today, a Claude Code release included some internal source code. No sensitive customer data or credentials were involved or exposed."

Business Insider - The leaked code was related to Claude Code itself, not the underlying AI models.

Business Insider - An X post featuring a screenshot of the leaked internal source code for Claude Code had accumulated 26 million views as of Tuesday evening.

CNBC - Anthropic leaked parts of the internal source code for its AI coding assistant, Claude Code.

Inc - The leaked code revealed the strategy behind Claude Code, which is valued at $2.5 billion.

WHERE:

CNBC - The leak pertains to Anthropic's Claude Code, which is a product of the company.

WHEN:

Inc - The leak occurred during the release of Claude Code's 2.1.88 update on March 31st, 2026.

Engadget - The leak occurred during the release of Claude Code's 2.1.88 update on March 31st, 2026.

Inc - The leak occurred on March 31st, 2026, and the confirmation was made on March 31st, 2026.

Business Insider - The leak occurred during a release on March 31st, 2026.

CNBC - The leak was confirmed on Tuesday, March 31st, 2026.

WHY:

Inc - The leak follows a public breakup between Anthropic and the Pentagon in February 2026, where the Defense Department opted to partner with OpenAI instead.

Inc - The leak occurred after a public breakup between Anthropic and the Pentagon in February 2026, where the Defense Department opted to partner with OpenAI instead.

Inc - The leak raises security questions about Anthropic, which has positioned itself as focused on AI safety.

Business Insider - The leak raises security questions about Anthropic, which has positioned itself as focused on AI safety.

CNBC - The leak was described as "a release packaging issue caused by human error, not a security breach."

HOW:

Inc - Anthropic is implementing measures to prevent similar leaks in the future.

Business Insider - Anthropic is rolling out measures to prevent similar leaks from occurring in the future.

CNBC - The leak occurred due to a mistake in the release packaging process.

(UPDATED April 1, 2026 19:20 UTC)

A significant leak has revealed details about Claude Code, one of the fastest-growing products in AI, following its 2.1.88 update released on Tuesday. Anthropic confirmed that the leak was due to a release packaging issue caused by human error, not a security breach, but the code was uploaded to a public GitHub repository and copied over 50,000 times before they could intervene. Among the features discovered in the leak were a "Proactive mode" allowing the AI to function without user prompts, a potential crypto-based payment system for autonomous transactions, and a Tamagotchi-like virtual companion that interacts with users' coding.

(UPDATED April 1, 2026 17:18 UTC)

A significant leak occurred during the release of Claude Code's 2.1.88 update on Tuesday, when a codebase was uploaded to a public GitHub repository and copied over 50,000 times. This leak exposed more than 512,000 lines of code and included 2,000 TypeScript files, revealing features like a "Proactive mode" that allows the AI to operate without user prompts, as well as a potential crypto-based payment system for autonomous transactions. Additionally, the leaked code mentioned a Tamagotchi-like virtual companion that would react to users' coding activities.

(UPDATED April 1, 2026 15:20 UTC)

A recent leak has unveiled the "Kairos" updates for Claude Code, introducing features that enable Claude to operate in the background and a new "dream mode" for memory consolidation. Following this leak, Anthropic's Claude chatbot saw a significant spike in downloads, briefly claiming the top spot in the US Apple App Store. This development comes after a public split between Anthropic and the Pentagon in February, when the Defense Department decided to collaborate with OpenAI instead.

(UPDATED April 1, 2026 13:22 UTC)

Anthropic's CEO, Dario Amodei, confirmed that the source code for their AI product, Claude, was leaked on Tuesday, giving competitors valuable insights into their development strategy. This incident raises significant security concerns for Anthropic, a company that has emphasized its commitment to AI safety. In response, Anthropic is now implementing measures to prevent similar leaks from occurring in the future.

(UPDATED April 1, 2026 09:20 UTC)

on Tuesday, a leak of internal source code related to Claude Code raised security concerns for Anthropic, a company that emphasizes AI safety. A spokesperson confirmed that no sensitive customer data or credentials were exposed, and a post featuring a screenshot of the leaked code has garnered 26 million views as of Tuesday evening. In response to the incident, Anthropic is implementing measures to prevent similar leaks in the future.

(UPDATED March 31, 2026 23:09 UTC)

on Tuesday, Anthropic confirmed the leak of parts of its internal source code for the AI coding assistant, Claude Code, which is valued at $2.5 billion. The company described the incident as a release packaging issue caused by human error, clarifying that it was not a security breach. This leak revealed the strategy behind Claude Code, raising concerns about the implications for the company's operations.