Open weights, closed risks: the fight over frontier AI's future
Anthropic's report calling Z.ai's downloadable GLM-5.3 a major cyberattack risk has reignited the debate over whether frontier AI models should be open or closed.
Topic
Topic
Anthropic's report calling Z.ai's downloadable GLM-5.3 a major cyberattack risk has reignited the debate over whether frontier AI models should be open or closed.