Open models have real benefits. Do they outweigh the risks?
MIT Technology Review weighs the case for and against open AI models. The debate was revived when Anthropic published a report last month arguing that the GLM-5.3 model poses a major cyberattack risk.