OpenAI has requested that the administration of U.S. President Donald Trump assist in protecting artificial intelligence (AI) companies from an increasing number of proposed state-level regulations, provided these companies voluntarily make their models available to the federal government.
اضافة اعلان
In a 15-page document containing policy proposals released on Thursday, the company behind the ChatGPT model stated that hundreds of AI-related bills currently under consideration across the United States could undermine the nation's technological progress, especially as it faces growing competition from China.
OpenAI explained that the U.S. administration should consider offering certain exemptions to both large and small AI companies from state-level regulations—if such regulations are enacted—if the companies agree to grant the government optional access to their models.
This recommendation was one of several proposals included in OpenAI's response to a request for public input from the White House, issued by the Office of Science and Technology Policy in February, as the administration works on crafting new policies to ensure the U.S. remains a leader in the AI sector.
Trump's AI Policy
Earlier, President Donald Trump canceled the comprehensive executive order on AI issued by the Biden administration and tasked the Office of Science and Technology Policy with developing an AI action plan by July.
To date, there is a noticeable absence of federal legislation regulating the AI sector. The Trump administration had generally indicated its intention to follow a non-interventionist approach in regulating this technology. However, several states are currently considering new measures related to everything from deep fakes to bias in AI systems.
Chris Lehan, Vice President of Global Affairs at OpenAI, stated in an interview that the U.S. Institute for AI Safety, a key government body focused on AI, could serve as the primary point of contact between the federal government and the private sector.
Legal Protection for Companies
If companies voluntarily cooperate with the U.S. Institute for AI Safety to review their models, the government could provide them with "legal protection, including exemptions from state-level regulations focused on the security of advanced models," according to the proposal.
Lehan said, "Part of the incentive to do so should be not having to deal with state laws that won't be as high quality as federal laws."
In its policy recommendations, OpenAI also reiterated its call for the government to take steps to support investments in AI infrastructure. Additionally, it called for reforms to copyright laws, emphasizing that the principle of fair use in the U.S. is crucial for maintaining leadership in AI.
The Threat of China’s AI Dominance
OpenAI and other AI development companies have faced several lawsuits related to copyright over the data used to train their models. The company stated, "If developers in China have unrestricted access to data, while U.S. companies are denied fair access to data, the AI race could effectively be over."
OpenAI also suggested that AI companies should have the right to access data held by the government, which could include healthcare information, according to Lehan.
The company stated that this information would help "advance AI development" and could become "critical if changes in copyright laws lead to restrictions on U.S. companies' access to training data." - Agencies
