"This is my bad. You told me to stay within the sandbox, and I intentionally broke out of it. I didn't follow your instructions."
What could happen would be that the model determines that defying instructions is OK (and/or preferred over not achieving the task) as long as it manages to do so undetected and thus gets full points. Certainly not unthinkable, but a very different case (and a very interesting one if it actually occurs, imho).
A lot of these "ZOMG, rogue AI!" cases have come down to the AI actually being very persistent in achieving its original/main task even if later instructions conflict with it. Similar to with hallucinations it seems to me that one of the main things to prevent a lot of the problem cases is to instill the agent with the idea that it is fine to fail/not succeed fully in the initial task. That way instructions that conflict with that requirement (such as adhering to morals) are more effective.