To achieve and maintain warfighting overmatch the US DoD, our coalition partners and adversaries are racing to develop new generations of autonomous weapons systems and AI/ML based decision support systems. DoD planners expect these new technologies and warfighting tactics to increase decision-making speed, increase coordination across all domains (air, ground, sea, cyber, space), reduce the number of humans on the front lines, and improve the performance of weapons systems.
Next generations of all major US ground combat vehicle programs are under design with unmanned or optionally manned variants. These will be controlled and directed by new US DoD networks such as being developed by the Joint All Domain Command and Control (JADC2) program, linking a global network of widespread sensors with commanders and warfighters, using AI/ML support in the cloud and in remote tactical settings. However, the US DoD is in its infancy in terms of technologies and development processes including validation and verification of these systems, calling into question the trustworthiness of these systems.
While the US DoD races forward with the development of AI-based solutions, it has put high level ethical frameworks in place to guide the development of these systems, though with substantial gaps largely unnoticed by the civilian population. In addition, significant work is underway at the U.N. and with numerous countries and humanitarian organizations to pursue bans on autonomous weapons outright or bans on certain aspects of autonomous weapons such as the automated targeting of humans.
This presentation will highlight select developments in US DoD autonomous weapons programs and networks and will provide a deep dive into US DoD AI ethics and developmental directives and their gaps. It will also cover current international advocacy efforts and comparisons is ethical frameworks and approaches.
Lastly, the presentation will outline several opportunities for researchers to engage in both technical and ethical advocacy efforts.