The authors define a Killer AI as a system employing artificial intelligence techniques that, either by design or by circumstance, directly lead to physical harm or death.

• This definition distinguishes between physical AI systems and virtual AI systems, with the intent of specifying how the latter can also be directly responsible for harm and death.
• AI systems that were previously incapable of being considered directly responsible for harm can now therefore be assessed more rigorously.
• To that end, the authors propose a framework that weighs the wellbeing of many people over the wellbeing of one or few people and on the basis that injuries not resulting in death can be recovered from more easily than those that do and are thus less severe.

Ensuring the Safety of this Emergent Technology

There has been no previous attempt to define the Killer AI phenomenon, even as the technology underlying it continue to become increasingly prevalent in our daily lives. The definition should help encourage further examination of AI and its risks—more comprehensive legislation, more appropriate regulation, more nuanced ethical discussion, and more research on the potential ubiquity of AI systems and how to minimize any harm they cause to the greatest extent possible.

For the full article, see  https://www.mercatus.org/research/defining-killer-ai