By Tom Leithauser
Leading AI (artificial intelligence) developers on November 9 pledged to address “dangerous misuse” of their frontier models for cyber attacks and other “risks arising from growing autonomy or model misalignment, including loss o ...