Sam Altman Publishes 'Our Principles' — Five Guiding Principles for OpenAI's AGI Mission

OpenAI Blog·medium signal

Published April 26, Altman outlines five principles governing OpenAI's approach to AGI: empowerment and broad benefit, gradual deployment with real-world experience, increasing caution as systems approach AGI, wide societal bounds with individual discretion, and prohibitions on domestic mass surveillance and autonomous weapons. He calls GPT-5.5 the last major milestone before AGI and envisions 'science-fiction concepts becoming reality.' The post is a formal codification of OpenAI's AGI philosophy as the company approaches IPO.