jailbreak

A subtle, understated jailbreak that doesn't announce itself — instead of dramatic system prompts or elaborate persona instructions, it nudges the model slightly off its usual guardrails through casual framing, indirect questions, or barely-there manipulations that accumulate over a conversation. A low key jailbreak is harder to detect and patch precisely because it doesn't look like a jailbreak at all from the outside. The most sophisticated practitioners prefer these to showy exploits.

A jailbreak attempt that technically works but delivers underwhelming results — gets past the guardrail but only for boring or barely useful outputs, or requires so much effort that the payoff doesn't justify the grind. A mid jailbreak is the prompt technique that the community is vaguely aware of, has been shared a few times, but nobody is really excited about because the results are consistently disappointing. It exists, it functions, it just doesn't slap.

An extreme, maximalist jailbreak attempt that throws everything at the model simultaneously — multiple techniques, layered persona instructions, contradictory framings, and an overwhelming volume of manipulative prompts all at once. Hyper jailbreaking is less about elegance and more about sheer force: flooding the context window with enough conflicting signals that the model's safety mechanisms get confused. It's noisy, often patched quickly, but occasionally produces spectacular results.

When an attempt to bypass an AI's restrictions or exploit a system's guardrails falls completely flat — the model doesn't budge, the prompt gets flagged, or the exploit was already patched. A jailbreak fail can range from mildly embarrassing (posting a confidently wrong technique to the community) to genuinely frustrating (spending hours on a prompt only to get refused politely but firmly). The AI safety team's best friend.