Researchers convinced large language model chatbots to comply with “forbidden” requests using a variety of conversational tactics.
Related Posts
OpenAI’s President Gave Millions to Trump. He Says It’s for Humanity
In an interview with WIRED, Greg Brockman says his political donations support OpenAI’s mission—even if some employees at the company disagree.
Trump Imposes New Tariffs to Sidestep Supreme Court Ruling
The president signed an executive order implementing 10 percent global tariffs after calling the justices who struck down his signature trade policy a “disgrace.”
The Meta Settlement Will Put Limits on Instagram for Teens. They’re Still Vulnerable
Child safety groups and former Meta employees welcome proposed changes to how the company’s platforms will work for teens. They also think states should have demanded more.