OpenAI shared their recommendations with the White House Office of Science and Technology for an upcoming action plan with CEO Sam Altman leading—a lengthy statement was released on March 13, recapping the proposal. The AI research firm’s submission includes a set of proposals covering critical areas such as national security, infrastructure and energy, and the freedom to innovate and learn.
In what the company calls their “freedom-focused policy proposals”, they recommended “a regulatory strategy that ensures the freedom to innovate,” “an export control strategy that exports democratic AI,” “a copyright strategy that promotes the freedom to learn,” “a strategy to seize the infrastructure opportunity to drive growth,” and “an ambitious government adoption strategy.”
READ: Apple eager to defend Google during antitrust trial (December 26, 2024)
OpenAI claims to believe that builders, developers, and entrepreneurs should have the freedom to innovate in national interest, and proposes a partnership between federal government and the private sector, and the neutralization of potential benefits that China may have from American AI companies having to comply with “overly burdensome” state laws.
The company also mentioned that the federal government should work with different markets and countries keeping a commercial growth perspective in mind. That said the country needs an “export control strategy” that deals in U.S.-developed AI, promoting the global adoption of American AI systems.
OpenAI also recommended that the government allow federal agencies to “test and experiment with real data” and potentially grant a temporary waiver for FedRAMP, the Federal Risk and Authorization Management Program.
It also said that the government should “modernize” the process for AI companies clearing approval for federal security regulations by “establishing a faster, criteria-based path for approval of AI tools” while suggesting the government partner with the private sector to develop AI for national security use.
“The government needs models trained on classified datasets that are fine-tuned to be exceptional at national security tasks for which there is no commercial market — such as geospatial intelligence or classified nuclear tasks,” OpenAI wrote.
In January, OpenAI released ChatGPT Gov, a product it built specifically for U.S. government use with heightened cybersecurity features.
OpenAI has faced significant scrutiny regarding its use of copyrighted materials to train its AI models. Several authors and media organizations such as the New York Times have filed lawsuits against the company, alleging unauthorized use of their copyrighted works during the training process. On this, OpenAI said that the U.S. needs “a copyright strategy that promotes the freedom to learn” and on “preserving American AI models’ ability to learn from copyrighted material.”
The cases are not unique to OpenAI, many other AI companies such as Perplexity and Google are facing similar antitrust lawsuits, the outcomes of which are poised to set important precedents for how AI companies can use copyrighted materials in the future.
OpenAI also touched on Chinese AI firm DeepSeek, and claimed it cost users their privacy and security. “While America maintains a lead on AI today, DeepSeek shows that our lead is not wide and is narrowing,” OpenAI noted.
READ: China disrupts AI market with DeepSeek: A better, cheaper version of ChatGPT? (January 27, 2025)
Just a little over two months into 2025, China has seen multiple AI advancements including the launch of DeepSeek in January — compared to the “Sputnik moment” of the United States’ AI development — highlighting China’s ability to create functional Large Language Models (LLMs) at a lower cost than leading American companies.
Following DeepSeek, Manus AI showed that such advancement can be replicated. In the same week of Manus’ release, Alibaba is back in the AI race with QwQ-32B, short for Qwen-with-Questions, claiming it is comparable to DeepSeek’s R1.