Bb5a31f81ClaimAnthropic's proposed legislation mandates independent safety evaluators registered with the Department of Commerce to report enumerated imminent risks within 72 hours of discoveryAgreeDisagree
Bb5a31f81ClaimAnthropic's proposed federal safety bill contained preemption clauses overriding state AI laws, including New York's RAISE Act, California's SB 53, and Illinois's SB 315AgreeDisagree
Bb5a31f81ClaimUnited States and Chinese government delegations held bilateral talks on artificial intelligence safety in Geneva in May 2024AgreeDisagree
Bb5a31f81ClaimRed team testing of Gemini 4 Argon demonstrated capability uplift over standard web search across two biological threat replication scenariosAgreeDisagree
Bb5a31f81ClaimTantum Collins served as Director for Technology and National Security on the National Security Council during the Biden administrationAgreeDisagree
Bb5a31f81ClaimGoogle published a safety and security report for Gemini 4 Argon spanning more than 200 pagesAgreeDisagree
Bb5a31f81ClaimBen Buchanan served as White House Special Advisor on Artificial Intelligence during the Biden administrationAgreeDisagree
Bb5a31f81ClaimAnthropic's draft legislation permits federal blocking of an AI model only if it presents a foreseeable, imminent risk exceeding $1 billion in damage or 50 serious casualtiesAgreeDisagree
Bb5a31f81ClaimBen Buchanan published an essay titled The US's AI Competition All Wrong in Foreign Affairs in the summer of 2020AgreeDisagree
Bb5a31f81ClaimAnthropic drafted and privately pitched federal AI safety legislation to United States lawmakersAgreeDisagree
Bb5a31f81ClaimAnthropic Head of Federal Affairs Anthony Cimino stated that preemption phrasing in Anthropic's draft legislation was imprecise and subject to revisionAgreeDisagree
Bb5a31f81ClaimGemini 4 Argon passed 14 of 15 stealth and situational awareness evaluation challenges administered by GoogleAgreeDisagree
Bb5a31f81ClaimDonald Trump announced openness to establishing artificial intelligence regulatory guardrails in cooperation with ChinaAgreeDisagree
Bb5a31f81ClaimThe Biden administration enacted export controls requiring export licenses for large-scale artificial intelligence chip purchases by countries outside close United States alliesAgreeDisagree
Bb5a31f81ClaimArtificial intelligence regulation was omitted from the pre-planned agenda of the Shenzhen trilateral summit between Donald Trump, Xi Jinping, and Vladimir PutinAgreeDisagree
Bb5a31f81ClaimDonald Trump stated that Donald Trump and Xi Jinping agreed that minimal guardrails on superintelligent artificial intelligence systems would be beneficialAgreeDisagree
Bb5a31f81ClaimThe Biden administration postponed the implementation date of its large-scale artificial intelligence chip export licensing regulations until May 2025AgreeDisagree
Bb5a31f81ClaimThe advanced artificial intelligence model Mythos was withheld from public release in early 2026 due to United States government capability vettingAgreeDisagree
Bb5a31f81ClaimCritics accused Anthropic of leading a race toward recursive self-improvement in artificial intelligenceAgreeDisagree
Bb5a31f81ClaimAnthropic established official policy red lines prohibiting its models from being used for domestic surveillance or controlling automated weapons systemsAgreeDisagree