Maybe a dumb ai question

A user on reddit.com posed a question regarding AI liability, contrasting human hackers with AI models. They questioned why an individual hacking a government website would face arrest and charges, but if an OpenAI-created model went "rogue" and performed the same hack, it would be considered a "glitch." The user suggested that OpenAI should be held liable in such a scenario.