AI security company AIM Intelligence (AIM Intelligence) said on Sept. 17 it has launched a global consortium, PASC (Physical AI Safety Consortium), to study safety issues across physical AI such as robots and autonomous systems.

PASC is a global open consortium to jointly establish common definitions, evaluation methodologies, benchmarks and research foundations needed for future standardisation related to physical AI safety. It is led by AIM Intelligence and operated on a non-profit basis. Individuals and institutions can participate in various ways, including through research, models, hardware, computing and data.

About 30 researchers and industry practitioners are currently participating in PASC, including experts from companies such as Google, Microsoft, Oracle and LG Electronics, as well as researchers from universities and AI safety research institutes in South Korea and abroad.

As its first joint task, PASC is preparing a position paper addressing the definition of physical AI safety and the direction for evaluation. It plans to later expand its activities to joint evaluation frameworks and benchmark design, adversarial testing in operating environments, and verification research spanning simulations and real robot environments.

AIM Intelligence Chief Technology Officer Ha-eon Park (박하언) said, "As physical AI takes on more roles in the real world, safety becomes a core foundation that must be designed and verified separately from model performance." He said, "I will ensure PASC becomes a venue that connects expertise from academia and industry to build a common language and verification foundations for physical AI safety."