Physical AI is now a reality.
Self-driving cars, care robots, and drones will move society alongside humans.
But the law still fails to define these beings.
"Physical AI and the Law" dissects the legal vacuum created by technology and raises the need to redesign responsibilities, rights, and governance.
Rather than suggesting how to control technology, we propose how to coexist with technology.
Artificial Intelligence Encyclopedia.
You can find the artificial intelligence knowledge you need at aiseries.oopy.io.

Physical AI refers to a system in which artificial intelligence technology goes beyond simple information processing or decision support, and moves and acts autonomously in a physical environment.
This is implemented in the form of artificial intelligence combined with physical devices such as robotic arms, self-driving cars, drones, and delivery robots, and has the ability to interact with the world in a human-like manner.
While general AI produces results such as text, images, and data analysis in virtual spaces, physical AI exists as an 'acting machine' that causes physical results through direct actions in the real world.
--- From "01_“What is Physical AI?”"

The concept of 'electronic personhood' began to attract global attention when it was officially mentioned in a 2017 European Parliament report.
This concept seeks to establish a legal framework for attribution of responsibility for damages or consequences caused by highly autonomous AI systems, particularly physical AI systems that perform physical actions.
The existing legal system assumes that humans or corporations are legal subjects, but the gap between AI's autonomy and impersonality raises problems that cannot be explained by a dualistic system.
--- From "06_“Legal Subjectivity of Physical AI: Discussion on the Introduction of ‘Electronic Personality’”

Currently, most physical AI devices in Korea are subject to general certification under existing laws, such as industrial machinery, medical devices, and electronic communication devices. However, these laws do not sufficiently reflect AI's unique judgment structure, update capabilities, and autonomy.
For example, care robots may not be subject to the Medical Devices Act, and delivery robots exist within the loopholes of the Road Traffic Act.
Therefore, an 'AI device-specific certification system' that takes into account the unique operating structure and risk characteristics of AI must be established, and this can be designed in parallel with the existing certification system.
New criteria are needed, including the ability to make autonomous decisions, the strength of interaction, and the ability to respond to unexpected scenarios.

--- From “08_“Physical AI Safety Certification and Permit System””