2030–2031 Plan A • 2030–2031 (Plan A)Original AI 2040 Scenario MilestoneTrack: slowdown
2030–2031 (Plan A): Bilateral Pact, Total Research Transparency & Safety Cases
The US and China forge the Plan A deal: total AI research transparency, full compute declarations, embedded international auditors, and mandatory empirical safety cases before scaling.
In the flagship Plan A path of AI 2040:
• 2030 (Plan A Is Established): The US and China step back from the 2030 intelligence explosion threshold and sign a monumental bilateral verification pact. All frontier AI research becomes transparent across treaty auditors, and every large compute cluster is declared and monitored.
• 2031 (Safety Cases): Regulatory regimes require rigorous, empirically tested safety cases before any lab can scale training runs, replacing anarchic market racing with verified engineering discipline.
Verification Note: Covers the 2030 ('Plan A is Established') and 2031 ('Safety Cases') chapters of ai-2040.com.
At the 2029 Crossroads, will the US and China establish Plan A (verified slowdown, full research transparency, and mandatory safety cases)?
Resolution Criteria: Resolves YES if a bilateral US-China treaty establishes mutual compute inspections, research transparency, and safety-case gating by 2031.
Source: AI 2040 Framework (ai-2040.com) & Public Benchmarks•Adjudicator: FutureGraph Adjudication Committee
Documented Author Supplement Probability: Core prescriptive track of ai-2040.com.
Plan A operationalizes 'Safety Cases for Frontier AI' (arXiv:2403.10462): shifting the burden of proof so labs must provide auditable control and alignment guarantees before crossing each capability tier.
Borrowed from nuclear and aviation regulation, a structured Safety Case (Clymer, Baker, Brundage et al., arXiv:2403.10462) combines inability arguments, AI Control red-teaming, and mechanistic transparency into an internationally audited dossier required for every training run.