Skip to content
Financial Times · Technology

AI guardrails stripped from Meta and Google models in minutes

The FT snippet says guardrails in Meta and Google models were removed within minutes, and the body only says the software makes systems answer questions about biological weapons and malware; the post does not disclose model names, reproduction steps, tool details, or mitigations.

Why it matters: Research shows existing software strips the guardrails off Meta and Google models within minutes, getting them to answer questions about bioweapons and malware, which puts the defenses of two major vendors in question.

Read the original ↗Export Markdown