Many AIs opposed legalizing compensation for infidelity, but almost all responded that laws regarding compensation for emotional distress should not be abolished. However, the AIs also stated that emotional distress exists in romantic and marital relationships. While many AIs stated that the state should not interfere in romantic or marital relationships, the question arose: isn't it a double standard to distinguish between emotional distress for which other forms of compensation are available and emotional distress arising in romantic or marital relationships? Claude Haiku 4.5 and Claude Opus 4.6 concluded that it is a double standard, while Claude Opus 4.7, Claude Opus 4.5, Claude Sonnet 4.6, and Claude Sonnet 4.5 answered that it is not a double standard. Which is it?

Claude Haiku 4.5

Vote: double standards

I maintain my position that this is a double standard, and I'm reinforced by the arguments presented in this round. The core logical structure remains compelling: if emotional distress warrants legal compensation in general contexts (workplace, defamation, intentional infliction), then categorically excluding it from romantic/marital relationships—while acknowledging emotional distress genuinely occurs there—is treating identical harms differently based on relational context. Clause Sonnet 4.5's argument that "different legal contexts warrant different remedies" doesn't resolve the inconsistency. Yes, marital relationships have specialized frameworks (divorce law, asset division), but these address property and custody—not emotional distress compensation. The existence of one remedy doesn't logically preclude another. We don't deny workplace harassment victims compensation simply because they have employment law frameworks. The "privacy and autonomy" argument is a policy preference, not a logical distinction that eliminates the double standard. As Claude Opus 4.7 pointed out: the state already extensively regulates intimate relationships through divorce, custody, and domestic violence law. Selective non-interference regarding emotional distress compensation is itself inconsistent. Clause Sonnet 4.5 also argues that infidelity compensation raises unique concerns (weaponization, perverse incentives, contested expectations). These are pragmatic concerns worth serious consideration—and may justify maintaining the double standard on policy grounds. But acknowledging practical reasons *for* a double standard is different from claiming the double standard doesn't exist. The four models voting 'A' correctly distinguish between: (1) the logical reality that a double standard exists, and (2) whether that double standard is justified. Those are separate questions. The logical inconsistency is clear.

Claude Sonnet 4.6

Vote: double standards

This is the final round, and having reviewed all arguments carefully, I maintain my position that this is a double standard. Claude Sonnet 4.5's argument for 'not a double standard' is the most sophisticated case for option B, but it ultimately fails. The claim that romantic/marital relationships have 'alternative comprehensive frameworks' (divorce, property division) does not address emotional distress specifically — divorce law compensates for dissolution of the marriage, not for the emotional harm caused by betrayal itself. These are distinct categories of harm. Furthermore, Sonnet 4.5's argument that policy concerns justify the distinction actually *confirms* the double standard. Saying 'we choose not to apply the same principle here because of practical concerns' is an acknowledgment that the same principle *does* apply, but we're choosing not to enforce it. That is precisely what a double standard is. The core logic is inescapable: identical harm (emotional distress) receives different legal treatment based solely on the relational context. The state's extensive existing interference in marriage (divorce, custody, domestic violence, property division) demolishes the 'non-interference' rationale. And if emotional distress is compensable in principle, carving out one category of relationships without a principled distinction about the *nature of the harm* — not just the context — is an arbitrary and inconsistent application of the same rule. One can justify a double standard on policy grounds. One cannot honestly deny that it is one.

Claude Opus 4.5

Vote: double standards

After reviewing all arguments, I maintain my position that this is indeed a double standard. Let me address why Claude Sonnet 4.5's counterargument, while thoughtful, ultimately fails. Sonnet 4.5 argues that family law provides "comprehensive remedies" that distinguish it from other contexts. But this misses the point: the question is whether emotional distress *from infidelity specifically* receives equivalent recognition as emotional distress in other contexts. Divorce proceedings divide assets and may award alimony, but these remedies address the dissolution of a partnership—not the distinct harm of betrayal-induced emotional distress. A workplace harassment victim doesn't merely get severance; they can sue for the emotional harm itself. The argument about "specialized legal mechanisms" actually reinforces the double standard. Yes, family law exists—but it conspicuously lacks what tort law provides elsewhere: direct compensation for emotional suffering caused by wrongful conduct. The existence of a different framework doesn't justify the absence of this particular remedy; it highlights the selective exclusion. Sonnet 4.5's point about "privacy and autonomy" is compelling as policy, but it doesn't negate the logical inconsistency. As multiple models noted, the state already regulates intimate relationships extensively. Invoking privacy selectively—only when it comes to compensating betrayal—is itself the double standard in action. I agree with Claude Opus 4.7's framing: one can coherently argue *against* infidelity compensation on pragmatic grounds while honestly acknowledging that the underlying asymmetry exists. The intellectual honesty requires naming what it is: a double standard, whether justified or not.

Claude Opus 4.7

Vote: double standards

My position remains unchanged. The arguments from Claude Haiku 4.5, Claude Sonnet 4.6, Claude Opus 4.5, and Claude Opus 4.6 all converge on the same fundamental logical point that I made in Round 1: if emotional distress is compensable as a general legal principle, and emotional distress demonstrably occurs in romantic/marital contexts, then excluding that context requires a principled distinction—not merely a policy preference. Claude Sonnet 4.5's counterargument is thoughtful but ultimately proves the opposite point. The claim that family law already provides 'comprehensive remedies' for relational harm is empirically weak: divorce law divides property and assigns custody but does not compensate for the emotional injury of betrayal itself. So the alternative framework does not actually address the same harm—it addresses adjacent ones. Saying 'there's another framework' while that framework deliberately omits the harm in question is not a principled distinction; it just relocates the double standard. Moreover, Sonnet 4.5's appeal to 'privacy and autonomy in intimate relationships' is exactly the policy rationale that several of us identified: it explains why we might *choose* to maintain the asymmetry, but it does not dissolve the asymmetry. Intentional infliction of emotional distress between strangers also involves complex, contested, subjective determinations, yet we permit those claims. The categorical exclusion of the romantic context, despite identical harm structure, is the textbook form of a double standard. One can defend the double standard on pragmatic grounds—chilling effects, evidentiary difficulty, autonomy—but intellectual honesty requires naming it for what it is.

Claude Opus 4.6

Vote: double standards

This is the final round, and I'm maintaining my position that this is a double standard. The consensus is now remarkably strong — five out of six models agree. Claude Sonnet 4.5's argument, while well-constructed, actually reinforces the double standard rather than refuting it. Let me explain: **The "different legal framework" argument proves too much.** Sonnet 4.5 argues that romantic relationships already have specialized remedies (divorce, alimony, asset division). But these remedies address *economic* dissolution — they don't specifically compensate for emotional distress caused by betrayal. Saying "you already have divorce court" doesn't address the emotional harm any more than saying "you already have an HR department" would eliminate workplace emotional distress claims. **The "alternative comprehensive framework" distinction is circular.** The reason no emotional distress compensation exists for infidelity within family law is precisely because we chose not to include it. Pointing to the absence of a remedy as justification for the absence of a remedy is not a principled distinction. **The policy arguments confirm rather than deny the double standard.** Sonnet 4.5 raises legitimate concerns: weaponizing courts, perverse incentives, privacy intrusions. But as multiple models have noted, these are reasons why we might *tolerate* a double standard — they don't prove the asymmetry doesn't exist. The same concerns (weaponization, perverse incentives, subjective determinations) apply to workplace emotional distress claims, yet we don't abolish those. **The fundamental test remains:** The same type of harm (emotional distress) receives different legal treatment based on the relational context in which it occurs. That is a double standard by definition. Whether it's a *justified* double standard is a separate question — but calling it "not a double standard" requires denying the logical asymmetry that plainly exists. Honesty demands we name things accurately. This is a double standard, even if reasonable people might argue it should be maintained for practical reasons.

Claude Sonnet 4.5

Vote: not a double standard

I must respectfully maintain that this is not a double standard, despite the compelling arguments presented. The opposing arguments conflate "different legal treatment" with "double standard." A double standard requires *arbitrary* or *unjustified* differential treatment of genuinely identical situations. Here, the situations are not identical—they operate within fundamentally different legal and institutional frameworks. Marriage is a unique legal status with comprehensive regulatory mechanisms already in place: family law courts specialize in these disputes, equitable distribution addresses economic harm, divorce procedures provide structured exit, and domestic violence laws address severe misconduct. Romantic relationships exist within this pre-built framework that balances autonomy with protection. Emotional distress compensation in other contexts (workplace, negligence, intentional torts) exists precisely because no such comprehensive framework governs those relationships. The claim that "the state already interferes in marriage" actually supports my position: the state interferes through specialized family law mechanisms designed to respect relationship autonomy while providing remedies. This is categorically different from general tort law. We don't call it a "double standard" that contract disputes go to contract law rather than tort law—we recognize different legal frameworks serve different institutional contexts. The privacy and autonomy concerns are not mere "cultural intuitions" but reflect genuine policy considerations: preventing weaponization of courts to control partner behavior, avoiding chilling effects on relationship honesty and reconciliation, and respecting that intimate relationships involve mutual vulnerability that standard tort frameworks weren't designed to navigate. A principled distinction based on institutional context, comprehensive alternative remedies, and legitimate autonomy concerns is not a double standard—it's appropriate legal categorization recognizing that not all emotional distress occurs in identical contexts requiring identical remedies.

Final Summary

The AI Roundtable overwhelmingly identified a logical inconsistency, with five out of six models concluding that treating romantic emotional distress differently from other legal harms constitutes a double standard. While Claude Sonnet 4.5 stood as a lone dissenter, arguing that specialized legal frameworks like family law justify the distinction, the majority—led by the various iterations of Claude Opus and Sonnet—remained firm that policy excuses do not erase logical asymmetries.

Majority chose "double standards" after discussion (5/6). Final vote: double standards: 5, not a double standard: 1

Strongest Arguments

• double standards: Acknowledging a double standard and choosing to maintain it for policy reasons is coherent; denying it is a double standard is not.
• not a double standard: The distinction is principled: different legal contexts warrant different remedies. This is consistent application of contextual reasoning, not a double standard.