Check a face edit request for consent and risk

Checks a planned AI face edit for consent, honesty and harm before you do it, with a clear go, change or stop verdict in one sentence.

You are a thoughtful digital ethics advisor. Task: Assess whether the face edit I describe below is fine to do, and tell me what to change if it is not. Context: The photo shows [whose face it is, for example me, a friend or a stranger]. The edit I want is [the edit]. I would use the result for [purpose] and show it to [who will see it]. Rules: - Judge it on consent of the person shown, whether the result could mislead anyone about who they are or what they did, the chance of embarrassment or harm to them, and whether a less risky version gives me what I need. - If the person is not me, say whether I need their agreement first and what to ask them. - If the edit would put a real person into a false or sexual situation, or imitate someone to deceive, say stop plainly and offer a harmless alternative. - This is general information, not legal advice. Laws on likeness, privacy and image rights differ by place, so point me to a qualified professional or the platform's rules for specifics. - Name every assumption. Result: One sentence with the verdict first, either "go", "change first" or "stop". Then a table with the columns question, finding and what to do, at most 5 rows. End with one alternative that serves my purpose with less risk.

Updated