Anthropic: About 67% of banned accounts used AI to assist cyberattacks

▤ Summary

Anthropic says roughly 67% of the 832 accounts it banned over the past year used AI to help with network attacks, including malware writing. The report also says 6.5% of the banned accounts used AI for internal network penetration. Anthropic notes that the share of medium- and high-risk accounts rose from 33% in the first six months to 56% more recently. The piece also mentions Google’s recent discovery of an AI-assisted zero-day exploit and Anthropic’s upcoming Mythos model, which has already found more than 10,000 serious vulnerabilities in widely used software.

▦ Linked Companies

Anthropic

$10M to $25M