From Behind the Desk into the Streets

As discussed before in class, robots and AI being used in police departments is nothing new behind the scenes. However, a new proposal now up for debate by the San Francisco Board of Supervisors has brought forward the question of whether robots should be allowed to use deadly force. This proposal, if passed, would allow robots the ability to use deadly force in the place of an officer. 

This has caused much concern from not only the public, but also from robotic companies. Boston Dynamics, one of the biggest robotics companies, has written an open letter to all police departments to not weaponize robots. Along with Boston Dynamics, legal experts such as Ryan Calo, a professor at University of Washington, raised some similar concerns. Stating that even though it limits the danger of human officers, the risk of it failing or overuse of the robots. Paul Scharre an expert on the topic has chimed in to ease some of the concerns. Stating that these robots operated by the police aren’t autonomous, but are operated by people. However, this has failed to ease many concerns as failure is still an issue.

The concern that arose about these potentially armed robots fall in line with our earlier discussions in class. The failure of machines are robots are a common occurrence as we saw with AI systems with facial identification. While Scharre mentioned these robots won’t be operated by AI, that doesn’t mean they will never be. If we allow robots to use deadly force in a similar manner as human officers and they continue to develop the technology to require the use of AI then it can be open to failure. The failure in this case wouldn’t be a wrong direction on google maps, but a tremendous loss of life.

In this vein, the legal questions about responsibility in the case of failure are yet to be answered. If we allow these lethal robots into service and they fail, who would be responsible? Would it be the operator, inventor, company, or service technician? In any case this would be the robots we have now without the AI, if we introduced AI into the mix it would expand the suspect pool. As of right now, San Francisco is trying to settle a debate that was not asked for nor one which has all the questions asked or answered.