Anthropic warns September 2 2026 malware drained Claude AI credits

Technology Oct 04, 2026

Heads up if you use Claude AI: Anthropic just warned that a malware attack hit some users on September 2, 2026.

Hackers used info-stealing malware to mess with user sessions and drain AI usage credits.

To keep things safe, Anthropic logged out affected accounts and wiped their payment info.

Anthropic refunding unauthorized charges

The attack mainly targeted Windows PCs and some Macs. Phones are in the clear.

Malware types like Vidar, Lumma, StealC, RedLine (on Windows), and Atomic Stealer (on Macs) were involved.

Anthropic is refunding unauthorized charges but says they're not necessarily responsible if you got malware from sketchy downloads like cracked software.

Their advice? Delete suspicious apps and run a full malware scan to keep your info safe.