Support for AI Safety Policies

Although respondents gave yes-or-no answers, their rationales often indicate that their support is conditional and their opposition is not absolute. “I support the policy only if…” is a common sentiment, as is “I don’t support the policy as written, but would if…” The qualifications typically concern acceptable costs versus benefits, the competence of potential regulators, international participation, and the alternative against which enactment is judged.

Opposition is the majority view, but much of it is centered on this specific version of preemption — which sets a ceiling but no floor — rather than the general concept: “I would change this answer if the policy were paired with a substantive federal standard”; “If it came bundled with a genuine federal standard I’d flip to yes, but as defined, no.” Among the substantial minority who support it, many cite the same concern when qualifying that support: “I support preemption if tied to strong federal AI security and control legislation.”

Where opposition is unconditional, the focus tends to be the value of state-level experimentation: “The individual state laws can continue to act as both incubators of new ideas and also as warning systems.” Unconditional support tends to focus on the value of uniform rules: “Better to have one set of federal rules than fifty different state ones.”

Supporters and opponents alike consider how much progress should be delayed, if at all, and which activities to constrain, rather than simply whether to back a slowdown. Many supporters acknowledge it could carry real costs, and many opponents accept it could buy useful time for safety work. There is a near-consensus that a bilateral version is preferable to a unilateral one.

US only. Opponents nearly always argue a unilateral cap would “allow China to get ahead in the race, in which case the risk from AI would likely increase drastically.” Others worry “domestic researchers [will] lose capacity for beneficial and defensive work.” Yet many opponents also express concern about the current pace: “A unilateral slowdown is not ideal in the long-term…But insofar as we're entering an exceptionally risky period of AI development right now (i.e., on the cusp of RSI), and the US is driving most of that risk, I think there are situations where a unilateral slowdown would be beneficial.” A recurring affirmative argument is that it preserves stronger options rather than working alone: “The risk reduction is small, but it could serve as precedent to make future more impactful policies more likely.”

US and China. Support is broad but typically conditioned on verification: “Only if it could be done in a manner that was reciprocal, verifiable, and narrowly tailored to focus on genuinely dangerous frontier capabilities.” Opponents tended to signal that their rejection was driven by verification concerns rather than the general merits: “[This policy] is inadvisable because it cannot be adequately verified.”

Many favor an independent check before dangerous models become widely available, yet many also condition support on the quality of the reviewer: “Do I want pre-authorization if unqualified judges are exercising very subjective judgement? No.”

US only. Supporters emphasize that their support depends on the nature of the authorizing entity: “[Only] provided the reviewing bodies possess genuine technical competence and authority to block dangerous deployments”; “[Only] provided the process is based on clear capability and risk thresholds.” Opponents point to the same variable: “[It] relies on federal government being a competent judge”; “[It risks] political bias leakage into model vetting.” As one writes, “I desperately want such a policy to be implemented. I just don’t think the quality is likely to be that good.”

International. Most who support the US version express even more support for an international regime, usually because it “would reduce the incentive for companies to move development to jurisdictions with weaker rules.” Skepticism is still common: “I don’t fully trust how it would work in practice. But I’d accept it as part of the wider package.” Several opponents support international cooperation but not the entity described: “International coordination is valuable, but [shared] standards and reciprocal verification are more workable than a centralized body with sweeping authorization power.” Unconditional opposition often centers on concentration of power: “Pre-release authorization risks entrenching a small number of approved developers, reducing model diversity and concentrating correlated failures.”

Responses to this proposal are particularly polarized, but much of the polarization stems from the policy details rather than the principle. Supporters emphasize that it “gives the AI companies the financial incentives to create safe products [and] forces them to focus on real-world outcomes instead of jumping through regulatory hoops”; “is the only policy that creates no new agency, but instead can be managed by courts and insurers”; and “conscripts private capital into risk assessment, pricing danger that regulators are too slow to see.”

A substantial share of the opposition nods toward those effects but objects that “this survey’s version allows damages for expected unrealized harms and mandates insurance,” adding, “that could become effectively prohibitory, so I would vote ‘No’ on this formulation.” Others write: “No to the full package as written.”; “Too harsh on AI providers, would completely destroy them.” Some supporters doubt it would be enough: “Strict liability I think is probably good, but I don’t think it actually guards against catastrophic risk - at that point, who cares if the company goes under?”

Many reject the full bundle because it includes a single policy they refuse to support: “[Strict liability] is a disqualifier”; “No, because I disagree with [international authorization].” Rather than rejecting it outright, some opponents describe the package they would accept: “I would support a revised bundle with workable financial assurance and proportionate, contestable remedies.”; “I would go ‘Yes’ on a bundle with [pre-authorization] and [strict liability], but not with the inclusion of a developmental slow down.” One respondent who answers No adds, “But…I would take this package if it was [the only thing] on the table.”

Support for the full bundle often reflects acceptance of an imperfect compromise. One respondent distrusts international authorization but would accept it if it helped secure “a joint slowdown and stronger accountability.” Another would accept the bundle “although I think strict liability is slightly harmful.”