Skip to content
Submit incident
AII-1000-49093
ConfirmedA court, a regulator or the organization itself has stated the core facts, or the record cites a court or regulator source.

Infostealer Malware Hijacks Claude AI Sessions and Drains Paid Usage

Infostealer malware on user computers stole active Claude login sessions, allowing attackers to drain paid usage quotas without user knowledge. Anthropic detected the activity, identified six malware families, and revoked sessions while refunding unauthorized charges, though the company noted the malware itself remains a threat if not removed from affected devices.

31 August 2026 Β· Anthropic Β· Security and cyberattacks
31 August 2026
AII-1000-490932 sources ↓
Quick facts
Category
Security and cyberattacks
Country
United States
Source
Security Affairs
Date
31 August 2026
Drafted by
Raidu Incident Detection

What happened

Infostealer malware on user computers stole active Claude login sessions without requiring passwords or bypassing multi-factor authentication, allowing attackers to access paid accounts.

Anthropic identified multiple infostealer families affecting Windows and macOS, including Vidar, LummaC2, StealC, RedLine, Acreed on Windows, and Atomic Stealer (AMOS) on macOS.

Reported impact

Attackers used stolen sessions to drain paid usage quotas without user knowledge, with some users reporting their usage limits refilling and then depleting while they were not using Claude.

Anthropic signed affected users out of Claude, removed saved payment cards, and refunded unauthorized charges.

Index analysis

The malware is general-purpose software unrelated to Claude itself, typically arriving through unofficial downloads or malicious apps, and collecting saved passwords, login cookies, and credentials from multiple services.

Infostealer malware bypasses traditional security measures by stealing authenticated browser sessions rather than targeting passwords or multi-factor authentication systems.

Governance controls

Anthropic detected the suspicious activity through its systems and revoked affected sessions.

Anthropic identified the malware families responsible and issued notifications to impacted users, though the company noted that revoking sessions alone is insufficient if malware remains on the user's device.

Classification

Category
Security and cyberattacks
Industry
Digital security
Country
United States
Harm status
Actual harm
Intent
Intentional
Timing
Post-deployment
Harm types
Economic or property
Harmed
Consumers
Failure modes
Data leak
AI systems
Claude

Sources and evidence

This record was researched and written by the Index, based on the following original reporting.

Security Affairs
Infostealers Are Hijacking Claude Sessions and Draining Subscriptions
31 August 2026
TimesNow
Anthropic Allegedly Warns Some Claude Users About Malware That Could Steal Their Data
31 August 2026