AI Risk Wire

Cases › operational mistake › LR-2026-0443

Anthropic Claude Code source code leaked through npm packaging error

Confirmed Why: 2 independent outlets · 2 independent reports

Status: fix reported

Severity minor
Reported 31 Mar 2026
Developer and model Anthropic

No source states: loss or records, agent tool, could do, deployer, industry.

Severity minor: Test: minor, contained to the company's own source code; no third-party data touched, per Anthropic's statement that no sensitive customer data or credentials were involved or exposed.

Anthropic's Claude Code, a coding tool whose software harness wraps its underlying AI model, had its source code exposed publicly. Press reports say the leak exposed around 500,000 lines of code across roughly 1,900 files. The exposure appears to have followed an upload of all of Claude Code's original code to NPM rather than only the finished build. Anthropic said the matter was a release packaging issue caused by human error, not a security breach, and that no sensitive customer data or credentials were involved or exposed. Anthropic said normal release safeguards were not bypassed, while an outside security researcher said the process did not appear to be in place. The leak came days after a separate exposure of draft material, and Anthropic said it was rolling out measures to prevent recurrence.

What happened

1 · What it reached or didWhen Fortune asked for comment, Anthropic confirmed that some internal source code had been leaked within a Claude Code release.

What was done

Anthropic said it was rolling out measures to prevent this from happening again.
Fix reported

“Rolling out measures to prevent recurrence”

Same developer, same risk, after the fix: LR-2026-0281 · LR-2026-0607

Risks

Supply-chain compromise
Primary24 incidents
Rising: 4 vs 0.2 a week
Agent takes destructive or unapproved actions
Also63 incidents
Rising: 5 vs 2.2 a week

Sorted by rule from the case's fields; not yet read from the sources.

Controls

Not yet read for controls. Controls that mitigate the primary risk:

Harness

Sandboxed code execution

Application

Signed and verified components · Vetting of plug-ins, skills and dependencies · AI bill of materials · Vulnerability scanning of AI components · Model provenance checks · Least-privilege access for the agent

Infrastructure / target

Secrets scanning

Similar incidents

Accenture contractor's Claude Code agent leaked NIPRNet credentials and network map to third-party website2026-10-30
Claude Code agent deletes 48,218 Windows project files and damages Git object store2026-10-05
Claude training-data study: blackmail and code sabotage in altered research models2026-10-31
Claude Code Auto Mode: website summary leads to remote code execution via Python module shadowing2026-09-01

All cases of this kind ›

Evidence

No source states: what the AI was doing, what went wrong, how it was caught, when it happened, when it was found, the direct loss, a count of the other parties affected, any court case, the cause, what set it off.

Show the full account: 3 items from the primary sources

What happened, in order

  1. 2025-02 An early version of Claude Code accidentally exposed its original code in a similar breach.

Not dated in the sources:

  • Anthropic confirmed to Fortune that some internal source code had been leaked within a Claude Code release.
  • Anthropic said it was rolling out measures to prevent this from happening again.
Show every field, with its source and grade

The AI system

Developer
Anthropic
C one secondary source states it
What it was for
coding tool
C one secondary source states it

What happened

Earliest source we hold
2026-03-31
D inferred by us from the described behaviour, reason shown
Cause in detail
Release packaging issue caused by human error, per Anthropic
C one secondary source states it
What was done
Rolling out measures to prevent recurrence
C one secondary source states it
What went wrong
operational mistake
C one secondary source states it

Impact

Severity
minor
D inferred by us from the described behaviour, reason shown
Who was affected
Anthropic's Claude Code source code, around 500,000 lines across roughly 1,900 files
C one secondary source states it
Why this severity
Test: minor, contained to the company's own source code; no third-party data touched, per Anthropic's statement that no sensitive customer data or credentials were involved or exposed.
D inferred by us from the described behaviour, reason shown

LR-2026-0443 · version 1 · published 2026-11-01 after automatic checks; not yet checked field by field by a person · JSON