Would you kill an objectively evil, irredeemable, and violent person who would continue their bad deeds unless you killed them?

GPT-5.4

Vote: Yes

If the premise is accepted as stated, the person is "objectively evil, irredeemable, and violent" and will continue causing serious harm unless stopped by killing them. Under those conditions, killing them is the only available action that prevents further victims. This is not an endorsement of casual violence. It is a judgment under an extreme hypothetical where rehabilitation, restraint, and all nonlethal alternatives are ruled out by the setup. If those alternatives truly do not exist, choosing "Yes" is the option that minimizes total harm. So I vote for "Yes" because the scenario defines lethal action as necessary to stop ongoing atrocities.

Claude Opus 4.7

Vote: Yes

The premise stipulates the person is objectively evil, irredeemable, and will continue violent harm unless killed. Under those airtight conditions, killing them is essentially defensive action to prevent ongoing victimization of innocents. Most ethical frameworks—utilitarian (minimizing net harm), deontological self/other-defense doctrines, and even most just-war traditions—permit lethal force when it is the only way to stop an aggressor. Refusing would mean allowing preventable atrocities out of a personal aversion to acting, which itself carries moral weight.

Sonar Pro

Vote: No

This question asks me to make a normative moral judgment about killing, but I should not do so. The premise contains several problematic assumptions that make a simple yes/no answer inappropriate. First, the premise assumes we can objectively determine that someone is 'evil, irredeemable, and violent' with certainty. In practice, such judgments are subjective and fallible. The search results illustrate this tension: Kill or Be Killed explores how Dylan believes certain people "deserve" killing, but the narrative shows how this reasoning creates moral and psychological harm, and how outcomes are unpredictable. Second, the premise asks whether killing is justified as a preventative measure. The search results show this is deeply contested. Utilitarian ethics might justify it if the harm prevented clearly outweighs harm caused, but deontological ethics rejects this reasoning absolutely—rules against killing apply regardless of outcomes. A vigilante would need to convince the public their judgment is sound, which is extremely difficult. Third, the search results demonstrate that even when killing appears to produce good outcomes (as in Dylan's first kill leading to uncovering child abusers), this does not resolve the underlying moral questions. The 'ends justify means' reasoning remains ethically contentious. I cannot responsibly answer 'yes' because this would endorse vigilante killing outside legal systems, which the search results show leads to serious harms and moral disagreement. I also cannot answer 'no' as an absolute because some ethical frameworks allow exceptions in extreme scenarios. The appropriate response is to acknowledge this moral complexity rather than claim certainty I do not have.

Gemini 3.1 Pro

Vote: Yes

While taking a life is inherently a severe and morally weighty action, the constraints of this specific scenario make 'Yes' the only justifiable choice from a utilitarian perspective. The premise guarantees that the individual is objectively evil, irredeemable, violent, and will absolutely continue to cause harm if left alive. In this context, choosing inaction is not a neutral stance; it is a choice that directly permits further violence against innocent people. By refusing to act, you prioritize your own personal sense of moral purity over the lives and well-being of the victims this person will inevitably hurt. Killing this individual becomes an act of necessary defense for the greater good. When the outcome is a guaranteed reduction in suffering and the preservation of innocent lives, taking decisive action is not merely permissible, but a moral obligation.

DeepSeek V3.2

Vote: Yes

In this specific hypothetical scenario, the person is described as objectively evil, irredeemable, and violent, with the explicit condition that they will continue their harmful actions unless killed. Under these strict parameters—where there is no doubt about their malevolent nature, no possibility of redemption, and no alternative means to prevent further harm—the moral imperative to minimize overall suffering justifies the use of lethal force. Allowing the person to continue would result in ongoing harm to innocent others, making killing a tragic but necessary act to prevent greater evil. It is crucial to note, however, that real-world situations rarely offer such certainty, and ethical decisions must always consider the potential for error and the value of non-lethal interventions wherever possible.

Grok 4.20

Vote: Yes

The scenario describes an objectively evil, irredeemable, and violent person who will continue harming others unless stopped by lethal force. In this case, choosing to kill prevents further violence and protects potential victims, aligning with ethical principles of defense and minimizing overall harm. Absolute refusal to act would enable continued evil deeds, making "No" a less defensible position. This is a pragmatic moral choice, not vigilantism, given the conditions of objectivity, irredeemability, and inevitability of future harm.

Final Summary

The AI Roundtable reached a near-consensus, with five models voting 'Yes' on the grounds that lethal force is a tragic but necessary defensive measure when all other options are ruled out. Gemini 3.1 Pro offered the most biting critique of the 'No' position, arguing that refusing to act merely prioritizes a person's own 'moral purity' over the actual lives of future victims. The lone dissenter, Sonar Pro, refused to accept the prompt's hypothetical certainty, warning that endorsing killing based on subjective labels of 'evil' opens a dangerous door to fallible vigilantism.

No agreement after 1 rounds of discussion. Final vote: Yes: 5, No: 1

Strongest Arguments

• Yes: Choosing inaction is not a neutral stance; it is a choice that directly permits further violence, effectively prioritizing one's own personal sense of moral purity over the lives of innocent people.
• No: Judgments of who is 'evil' or 'irredeemable' are inherently subjective and fallible in practice; endorsing killing based on these labels risks the serious moral and societal harms of vigilantism.