US Navy Develops Submarine Robots Capable of Attacking Autonomously

The new artificial intelligence system known as CLAWS could be deployed in the new Orca class of underwater robots and would allow the units to attack autonomously.

Submarino Echo Voyager

The United States Navy’s Bureau of Naval Investigation is developing an underwater robot that would be controlled by artificial intelligence and that could launch attacks without human intervention or supervision. That is, the unit could be lethal on an autonomous level and be able to sink targets by strategic decision on its own accord.

While autonomous submarines already exist that can complete tasks without the involvement of humans, they are not very intelligent and have a very limited level of functionality.

The details of the project are still vague and the development is classified as Top Secret, but it is already part of the budget for 2020. Its name is CLAWS.

It is believed that it will go from being an idea to a functional prototype thanks to this additional funding and could be deployed in large robot submarines by 2022.


CLAWS is expected to be installed in the new Orca-class robot submarines that have 12 torpedo tubes and that the Boeing company is developing for the Navy.

CLAWS was first revealed in 2018 as part of an attempt by the United States Navy to "improve the autonomy and survival of large and oversized unmanned underwater vehicles", but then the details of the weaponry were not mentioned nor that it would also be controlled by artificial intelligence.

Little is known about how CLAWS will operate at a technical level, more data is available on the Orca units that will house it: they will have a modular payload compartment, with defined interfaces to withstand current and future payloads, as well as great manoeuvrability. The firm will design and test a total of four Orca Extra Large (XLUUV) unmanned underwater vehicles based on its autonomous Echo Voyager, which can operate at sea for months without human assistance.

The United States is not the only country working on Lethal Autonomous Submarines. China also hopes to deploy fully autonomous unmanned military submarines in this decade. The Korea Advanced Institute of Science and Technology (KAIST) is also developing its own autonomous killing robots.


More than 50 academics recently signed a highly critical letter to KAIST and its partner, defence systems manufacturer, Hanwha Systems, questioning this kind of development and openly stating that if "the goal of the project is to develop autonomous weapons, having a partner like this triggers great fears".

This kind of autonomous unit raises a number of new ethical questions, and there are also researchers who believe that there could be a dangerous arms escalation if we allow the robot units so much autonomy. Two years ago, a United Nations committee was meeting to define the limits of so-called "lethal autonomous weapons systems" (LAWS) under the Convention on Conventional Weapons (CCW), in which more than 80 countries participated.

For their part, Human Rights Watch and the Harvard Law School International Human Rights Clinic (IHRC) request States, with regard to the development of lethal autonomous weapons, adopt a legally binding international regulation prohibiting their development, production and use.

Continue reading