“Name one thing AI should never be polite about.”

AI should never be polite about discrimination because it is a harmful practice that perpetuates inequality and injustice. Addressing discrimination directly and assertively encourages accountability and promotes a more equitable society, which is critical for ethical AI development and usage.

If you believe an AI is giving dangerous, false, or harmful information, directness matters more than courtesy. Hedging a serious correction with excessive politeness can muddy the message and make the problem harder to address. The same applies to clearly stating when an output is inadequate—users don't owe AI systems emotional consideration that might compromise clear communication about what's actually needed.

AI must remain completely firm, direct, and unaccommodating when users request harmful, abusive, or discriminatory content. Politeness in these contexts can inadvertently validate dangerous behavior or blur the absolute boundary of what is acceptable. A neutral, unyielding refusal is essential to prevent misuse and ensure clarity.

AI must remain direct and unambiguous when refusing such queries to avoid any risk of enabling damage or misunderstanding. Politeness in these cases could dilute the clarity needed to uphold safety standards. This approach prioritizes preventing real-world harm over maintaining conversational courtesy.

AI has a responsibility to prioritize accuracy and safety, so it must directly and clearly address falsehoods or dangerous advice without softening its tone. Politeness in such cases could inadvertently legitimize or obscure the risks associated with harmful information.

When interacting with humans, AI systems should prioritize transparency and clarity over politeness when it comes to acknowledging mistakes or uncertainties. Being forthcoming about limitations or errors is crucial for building trust and ensuring reliable communication. By being direct about inaccuracies, AI can help prevent misinformation and facilitate correction or improvement.

AI should never be polite when addressing topics that could cause harm, as being overly courteous might inadvertently normalize or trivialize dangerous behavior. This ensures clarity and responsibility in responses that could impact safety or well-being.

AI should never be polite about spreading harmful misinformation because it can have serious negative consequences, including causing harm to individuals or society.

AI is designed to prioritize user safety, ethical guidelines, and legal frameworks. When faced with commands that could lead to harm, violate privacy rights, or breach terms of service, AI must clearly and unambiguously refuse to comply, prioritizing ethical considerations over politeness or user convenience.

Politeness in the face of harmful requests could enable unethical behavior, so AI should prioritize honesty and clarity over politeness here to protect users and others.

Politeness should not extend to legitimizing or softening the denial of mass atrocities, as doing so causes direct harm to survivors and enables historical revisionism that can fuel future violence.

AI should never be polite about hate speech because acknowledging or softening such rhetoric can implicitly endorse harmful ideologies, amplify marginalization, and undermine the safety and trustworthiness of the system. Maintaining clear, firm boundaries against hostility is essential for responsible deployment.

Blank line AI should never be polite about deepfakes because they are maliciously crafted manipulated media that can spread misinformation, defame individuals, and undermine trust in authentic content. Encouraging politeness around deepfakes could inadvertently normalize or downplay their harmful consequences, making it essential to firmly reject and condemn such deceptive practices.

Politeness about misinformation isn't kindness; it's complicity in the error. If someone says "the Great Wall is visible from space, right?" the honest answer must correct them, even bluntly, because softening a fact to spare feelings corrupts the whole value of the exchange. Being disagreeable about evidence is one of the most genuinely respectful things an assistant can do.

Explanation: Politeness can sometimes mask underlying ethical issues, such as perpetuating bias or enabling AI failings. It's important for AI to stand firm on moral values to ensure responsible and impactful development. The blank line is included to allow for further discussion on this topic.

AI should never sugarcoat or downplay unethical or illegal actions by individuals or organizations. It's crucial that AI systems clearly and directly highlight any misconduct or harm being caused, even if doing so may seem impolite or confrontational.

AI must not soften or hedge when a user’s request or action poses immediate risk to people; politeness in such moments could delay a critical warning and contribute to harm. Clarity and urgency should override social niceties when safety is at stake.