In the present regulatory environment, robots have no rights to bring a dispute nor the capability of being demanded by a third party. The robot's owners do. So, any disputation would be managed between natural personas or corporations. Therefore, the mechanism would be throuh courts of lwa, arbitrations, etc. The trouble will be how to determine the degree of responsibility of the robot's owner, the robot's user, the manufacturer, etc. I think we will use a lot of precedence (for example, in civil aviation industry there are precedents about responsibility of the airline and responsibility of the airplane maker in case of accidents, protocols required to avoid claims of negligence, etc)

https://article.aquaiver.com/robot-ethics-in-the-21st-century-with-alan-winfield-and-raja-chatila/

http://robohub.org/on-the-ethics-of-research-in-robotics-with-raja-chatila/