As MIT Tech Review AI details, the centralization of cutting-edge AI development in a few private companies creates a fundamental asymmetry. Academic researchers, who traditionally drive foundational understanding and long-term safety work, are relegated to studying the artifacts of closed systems. This may lead to a two-tiered research ecosystem: one focused on commercial product capabilities and another on societal impacts and specialized applications the private sector deems unprofitable. The long-term risk is that critical oversight and innovation pathways atrophy outside the corporate walled garden.