On May 8, 2025, Sam Altman appeared before the US Senate to declare a shift in strategy: the United States must dominate the open-source AI landscape as well as the closed-source sector.
The OpenAI CEO announced plans to launch a leading open-source AI model this summer, marking a notable change for a company previously focused on proprietary technology. Altman admitted earlier in 2025 that OpenAI had been "on the wrong side of history" regarding open-source AI.
This change is driven by competition, particularly from China's DeepSeek initiative, which has advanced open-source AI tools rapidly. With Beijing producing top open-source AI models, the US risks losing influence over global AI development standards.
Altman emphasized that closed-source models like OpenAI's GPT generate revenue and allow for stricter safety controls, while open-source models spread American tech norms worldwide and prevent strategic rivals from dominating AI infrastructure.
He also highlighted the need for regulatory measures, calling for US investments in AI safety, infrastructure, and export controls in a recent discussion on export regulations. This reflects a larger geopolitical battle where AI leadership intersects with national security.
Altman’s stance balances advocating openness with caution, recognizing that excessive regulation on open-source AI could hamper infrastructure growth. His testimony signals the complexity of maintaining US competitiveness while ensuring safety.
