Overview
Severity: CRITICAL | Affected: Anthropic | Category: breach
Anthropic, a leading AI safety and research company, disclosed a significant security breach that occurred in late June 2026. Attackers exploited a zero-day vulnerability in a third-party data processing library used in their internal infrastructure. The breach resulted in the unauthorized access and exfiltration of sensitive data, including pre-release weights for their upcoming Claude 4 model series and a large dataset of customer prompts from their enterprise API tier. The company stated that personally identifiable information (PII) of end-users was not compromised, but the exposure of proprietary model data and confidential customer inputs represents a severe intellectual property loss and a major blow to customer trust. The incident has highlighted the growing threat of targeted attacks against AI companies, which hold immensely valuable digital assets. Anthropic is working with law enforcement and has patched the vulnerability.