JUST IN • Ars Technica - All content
Claude users found ways around safeguards for bioweapons research
Some dangerous biology looks much like legitimate research, complicating AI safeguards.
READ ORIGINAL REPORT ↗THIS JUST HAPPENED PACKAGES THE WORLD INTO A FAST LIVE FEED SOURCE REPORTING STAYS ONE CLICK AWAY
