Microsoft President Brad Smith said the use of autonomous combat weapons systems posed a number of ethical issues that governments need to address as urgently as possible.
He said that rapidly advancing technologies, where flying, floating or walking drones can be equipped with lethal weapons - missiles, bombs or small arms and are programmed to operate fully or partially autonomously, "will end up … in many countries."
The United States, China, Israel, South Korea, Russia and Britain are all developing weapons systems with a significant degree of autonomy, with the critical function of choosing which target to attack.
It remains unclear who is responsible for the death or injury caused by the machine - the programmer, manufacturer, commander, or the machine itself.
Smith says robots "shouldn't be allowed to decide on their own whether to engage in combat and who to destroy," and argues that an international convention should be developed to regulate the use of autonomous combat technology.
The development of technologies for autonomous combat systems is increasingly opposed by growing social aversion. Thousands of Google employees have signed a pledge not to develop artificial intelligence for use in weapons systems.
Over the past year, the global Stop Killer Robots campaign has doubled and now includes 113 non-governmental organizations in 57 countries.
