INTERNATIONAL POLICY FRAMEWORK: Regulating Artificial Intelligence in STEAM Discovery

This policy framework addresses the systemic risks of compute asymmetry and epistemic capture caused by AI corporate dominance over high-performance computing infrastructure in scientific discovery. To safeguard human comprehension and mathematical integrity, this paper proposes a binding, two-tier global governance architecture managed by a new proposed United Nations body: the International Board for Algorithmic and Mathematical Integrity (IBAMI). Enforcing a three-part technical audit on commercial entities whose resource allocations exceed standardized Academic Compute Thresholds, the framework mandates a 90-day public declaration rule and a 180-day execution freeze to protect academic work. Finally, it establishes cross-border enforcement protocols, including absolute academic nullification and judicial hardware disconnection.