The AI Futures Project released AI 2040: Plan A, a comprehensive scenario that recommends delaying the creation of superintelligence until 2040 rather than allowing it to emerge in 2030.

The scenario presents what the research group calls a "recommendation, not a prediction" — outlining what should happen rather than forecasting what will occur. The plan assumes decisive intervention by US and Chinese governments to slow AI development timelines.

Unlike previous high-level discussions focused on abstract concepts of "doom or utopia," AI 2040: Plan A works through implementation details step by step. The scenario builds on the group's earlier AI 2027 work, which explored shorter-term AI development trajectories.

Why the 10-year delay matters

The proposed timeline shift from 2030 to 2040 reflects growing concerns about rushed superintelligence development. The scenario assumes current AI progress would naturally lead to superintelligence by 2030 without regulatory intervention.

The AI Futures Project spent the past year developing the scenario, which they describe as "detailed and comprehensive." The full scenario is available at ai-2040.com for researchers and policymakers examining AI governance approaches.

The timing coincides with increased government attention to AI safety and development timelines. Both US and Chinese authorities have signaled interest in AI governance frameworks, though specific regulatory approaches remain under development.

The scenario joins a growing body of work examining AI development pathways and intervention points. Other research groups, including teams at Anthropic and independent safety researchers, have explored similar questions about optimal AI development timelines.

The AI Futures Project positions the work as actionable guidance rather than speculative fiction. The scenario's focus on government intervention reflects the group's assessment that private sector coordination alone may prove insufficient to manage superintelligence development timelines.

The project plans to use the scenario for policy discussions and research planning. The detailed implementation approach distinguishes it from broader AI safety discussions that often remain at conceptual levels.