Hackers reportedly used Claude and Grok to automate attacks on more than 400 financial targets in Latin America

Positive accounts highlighted the scale of AI agents automating attacks on 400+ banks, while negative replies condemned bypassed guardrails, absent human oversight, and the offensive edge over bank defenses.

Based on 85 sentiment-bearing replies from 84 accounts across 5 conversations.

The Information also reported that the attackers bypassed AI safeguards and stole sensitive data with little human involvement. Its public posts do not identify individual targets, date the attacks or specify what data was stolen. They give no count of successful intrusions.

Positive accounts highlighted the scale of AI agents automating attacks on 400+ banks, while negative replies condemned bypassed guardrails, absent human oversight, and the offensive edge over bank defenses.

Based on 85 sentiment-bearing replies from 84 accounts across 5 conversations.