Anthropic at least used to stop conversations cold when they reached the end of the context window, so it's entirely possible from a technical standpoint. That

Anthropic at least used to stop conversations cold when they reached the end of the context window, so it's entirely possible from a technical standpoint. That OpenAI chooses not to, and prefers to let the user continue on, increasing engagement, puts it on them.