AI pioneer Fei-Fei Li urges independent oversight of AI, arguing safety assessments must not rest solely with developers. She advocates for benchmarks set by governments, academia, and industry to ensure human control and balance technological benefits with safety. Her call for shared responsibility contrasts with recent claims by US President Trump dismissing existential AI risks as a scam. As AI capabilities grow, Li emphasizes that humanity must maintain control over its development and use, addressing concerns raised by other tech leaders about potential risks beyond human control.