Infostealer-Driven Claude Session Hijacking
First seen Aug 31, 2026 · Updated Aug 31, 2026
Anthropic has warned that infostealer malware infecting user PCs is exfiltrating active Claude session tokens, allowing attackers to hijack accounts and consume victims' paid usage. This represents a growing trend of infostealers specifically targeting AI service credentials and session cookies rather than just traditional banking or email accounts.
Technical Analysis
Commodity infostealer malware (families such as Redline, Lumma, or similar stealer-as-a-service tools commonly deployed via cracked software, malicious ads, or phishing) harvests browser-stored session tokens and cookies, including active authenticated sessions for Claude.ai. Because session tokens bypass MFA and password re-authentication, attackers can directly hijack live sessions without needing credentials, enabling unauthorized API/usage consumption and potential exposure of conversation history or connected data. This attack vector is particularly concerning for organizations using Claude as part of agentic workflows, RAG pipelines, or developer tool integrations, since a hijacked session could allow attackers to exfiltrate proprietary prompts, business logic embedded in system prompts, or connected data sources, and abuse API-linked automation under the victim's identity and billing.
Affected Systems
Claude.ai web sessions and associated browser session tokens/cookies on infected Windows/macOS endpoints; any system where Claude credentials or session tokens are cached in browser storage or synced credential managers
Indicators of Compromise
- No specific hashes, IPs, or domains disclosed in source reporting; organizations should monitor for anomalous Claude account usage spikes, logins from unrecognized geolocations/IPs, and endpoint infections by known infostealer families (Redline, Lumma, Vidar, RisePro)
Remediation Steps
- 1
Rotate and revoke sessions
Immediately log out of all active Claude sessions and reset account passwords; enable MFA where supported.
- 2
Endpoint scanning
Run endpoint detection and response (EDR) scans to identify and remove infostealer malware on affected devices.
- 3
Monitor usage logs
Review Claude account billing and usage logs for anomalous activity, unexpected API calls, or usage spikes indicating unauthorized access.
- 4
Restrict credential storage
Avoid storing session tokens/cookies in browsers without protection; use hardware-backed credential managers and enforce session timeout policies.
- 5
User awareness training
Educate employees about risks of downloading cracked software, suspicious browser extensions, and phishing lures that commonly deliver infostealers.
- 6
API key and integration audit
For organizations using Claude in agentic or automated pipelines, audit and rotate any API keys tied to compromised accounts and review connected integrations for unauthorized access.
Industries Most Exposed
Respond to this threat
Pro subscribers get a full AI-generated incident-response playbook for this threat — detection, containment, eradication, and recovery steps — plus an unlimited AI Threat Advisor for questions about your environment.