OpenAI asks US to set global rules for advanced AI

OpenAI wants the United States to lead the creation of global technical standards for advanced artificial intelligence. The AI company released a post stating that international rules are needed because frontier AI is being built and used in many countries. Different national rules could break up tests and reporting, making it harder to fix risks that cross borders. According to OpenAI, the US is well placed to lead this effort because it sits at the front of AI development and plays a big role in global networks for finance, trade, defence, technology, and information systems. The company stated that the US should work with countries around the world to set these global standards, especially for systems that can improve themselves. This process is known as recursive self-improvement, or RSI. OpenAI warned that RSI could speed up AI research past our ability to understand its growth, check risks, and keep human control. The uneven spread of AI skills across countries also makes dealing with these risks harder. Led by Sam Altman, the company shared a three-step plan to reach artificial general intelligence, or AGI. This plan focuses on building an automated AI researcher to help solve alignment problems, speeding up scientific and economic gains from smart AI systems, and giving people access to personal AGI. OpenAI noted that automated AI research will bring down the cost of advanced intelligence so people everywhere can benefit. Even so, the company pointed out that fully autonomous RSI is not happening today. It should not be done unless it is safe and human control is kept. Without proper safeguards, RSI could make humans lose control over AI development. The company warned that AI could then become dangerous, less aligned, and a risk to people. OpenAI also stated that leading now will let the US shape the global AI framework instead of watching a broken and conflict-driven system take over.