Trump Says US Rejects Any Scheme to Control AI

President Donald Trump told the UN General Assembly that the US rejects any “globalist scheme” to control artificial intelligence, while pledging close monitoring without stifling AI growth. The administration will also refer to AI as “super intelligence.” The remarks signal a potentially less restrictive US stance on AI development but provide no specific policy measures, timelines, or regulatory details.

Analysis

This is primarily a policy-risk signal rather than an earnings catalyst: the market should not capitalize it until it is translated into agency guidance, procurement rules, power-permitting policy, or export-control changes. A less restrictive domestic posture marginally supports the valuation premium of US AI platforms (MSFT, AMZN, GOOGL, META) and compute vendors (NVDA, AVGO), because their bottleneck is increasingly deployment speed—data-center interconnection, power availability, and enterprise adoption—rather than model capability. The near-term effect is likely limited given that antitrust, privacy, copyright, national-security and state-level constraints remain independently binding.

The more actionable second-order implication is geopolitical fragmentation, not deregulation. A US refusal to accommodate multilateral AI governance could accelerate sovereign-AI spending and parallel compliance stacks, benefiting infrastructure and cybersecurity vendors with government/regulated-industry exposure (ORCL, PLTR, PANW) while raising localization and compliance costs for globally distributed software vendors. Over 6-18 months, the risk is that tougher European or Asian rules become de facto product constraints for US platforms, reducing scale economies and increasing legal/operating expense; that outcome would favor firms able to sell private-cloud or on-premise AI deployments over pure consumer-model monetization.

Consensus may overread political language as a blanket green light for AI capex. The binding constraints remain grid capacity, returns on incremental data-center investment, and export restrictions on advanced compute; an aggressive deregulatory narrative could even worsen power-equipment inflation and lengthen build cycles. The thesis is falsified by concrete federal rules that impose model licensing or liability standards, broader semiconductor export controls, or hyperscaler earnings showing AI capex rising faster than cloud/AI revenue conversion.

Key Decisions for Investors

• No directional index-level AI trade solely on this speech; monitor for executable follow-through over the next 30-90 days, especially federal procurement directives, permitting actions, or export-control revisions.
• Maintain a 6-12 month quality tilt toward MSFT and AMZN versus high-multiple, pre-revenue AI software: these firms can monetize regulatory/compliance complexity through enterprise cloud distribution, while smaller application vendors face higher customer procurement friction. Reassess if cloud growth fails to stabilize or capex-to-revenue conversion deteriorates for two reporting periods.
• Watch-list pair for a confirmed sovereign-AI/regulatory-fragmentation trend: long ORCL or PLTR / short a broad software basket such as IGV, sized only after evidence of government or regulated-industry contract acceleration. Target a 10-15% relative move over 6-12 months; exit if bookings do not improve or federal procurement remains unchanged.
• Avoid adding to NVDA or power-infrastructure exposure purely on perceived deregulation. Add only if hyperscaler capex guidance is raised and data-center delivery schedules improve; a further extension of utility interconnection queues or export-control tightening would impair the expected demand realization despite supportive rhetoric.