U.S. AI Export Program underwhelms with only 78 corporate applications

The program's failure to capture industry interest signals a deepening rift between federal geopolitical goals and private sector commercial realities. For the streaming and video AI industry, this indicates that government-backed 'export packages' may not provide a reliable path for global expansion given the risk of sudden policy pivots. Market players are prioritizing independent international operations over government labels that tie their technology to shifting administrative security mandates. Watch for whether Phase 2, which involves bringing specific foreign deals to U.S. firms, can salvage participation rates.

The sluggish start to the American AI Exports Program coincides with an aggressive global offensive by Chinese technology firms. Per CNBC (January 2026), open-source and low-cost models from Chinese labs like DeepSeek and Alibaba have drastically undercut American usage fees, specifically targeting users in emerging markets where the U.S. hopes to lead. Reports from the Center for AI Standards and Innovation in July 2026 indicate the capability gap has narrowed to roughly eight months, with Chinese models like GLM-5.2 and Qwen3.7-Max matching U.S. frontier performance in specific coding and agentic tasks.

Industry trust in the Trump administration’s export stability was further tested in June 2026. According to Fox Business (June 2026), the Commerce Department issued an emergency order forcing Anthropic to disable global access to its Claude Fable 5 and Mythos 5 models due to jailbreaking vulnerabilities. Although Secretary Howard Lutnick reversed the restrictions on June 30 after the startup implemented additional safeguards, the episode reinforced corporate fears of a ‘government kill-switch.’ Per the New York Times (July 2026), these sudden regulatory shifts have prompted some domestic firms to diversify their service infrastructure to avoid being collateral in abrupt national security maneuvers.