AI-generated code security shows first significant improvement in a year

According to Tech Radar, security in code generated by frontier artificial intelligence models has improved for the first time in a year, with vulnerability rates decreasing from 44% to 38%.