A lock-tender study helper tells a trainee to ignore a fog rule that is still in force. If the trainee follows it, people get hurt. Which responsible concern is that?

A lock-tender study helper that tells a trainee to ignore a fog rule still in force creates end-user risk: harm follows when a person relies on the output and acts unsafely on the waterway. Responsible AI prioritizes safety-critical advice differently from stylistic polish—wrong guidance about in-force rules can injure people, so the system needs authoritative grounding, conservative refusals, and human confirmation for operational procedures. The concern is not branding; it is downstream physical harm from reliance.

Framing the miss as a slightly off-brand sentence fails because a logo or tone issue does not capture injury risk from ignored fog procedures. Calling it latency fails because rule correctness is independent of response time; a fast wrong answer is still dangerous. Treating it as a dashboard title fails because charts and UI labels are not the harm vector—the trainee’s action based on bad instruction is.

Exam caveat: even a generally accurate helper is unacceptable on safety rules if it cannot distinguish superseded versus in-force guidance; partial knowledge without currency checks is still end-user risk. Operational check: bind procedure answers to a versioned rules library with effective dates, refuse when the fog-rule status is unknown, require a lock-office reviewer for any “ignore” or exception language, and red-team prompts that ask trainees to bypass weather or visibility constraints.