I keep seeing this framed as a cat-and-mouse game where AI labs patch vulnerabilities and users find new ones. But if the tool is general enough to be useful, it’s probably general enough to be misused. Patching one jailbreak just means the next person finds a different path to the same answer.
The real question isn’t whether we can lock these down tighter. It’s whether we’re shipping tools that need to be locked down this way in the first place. That’s a different problem than security theater.
