Opponents of arming robots are pushing back, hard, against a supes policy that would give the San Francisco police the right to have the machines use deadly force—and there’s some indication they are making progress.
Sup. Gordon Mar, one of the eight who approved the policy, said today that’s he’s changed his mind and will vote No when the measure comes up for a second and final vote Tuesday.
A letter signed by 44 community and civil-liberties groups is urging the board to reject the policy. And Sup. Dean Preston has warned that the previous vote was illegal and asked that the measure be sent back to committee. Sup. Aaron Peskin, who supported the plan, told me tonight he would vote to send it back, suggesting that the supes will not sign off on this proposal.
The proposal, which came up at the last minute after months of hearings and discussions on police weapons, would allow senior SFPD officials to authorize a robot to carry a bomb that could be used to kill a person.
The cops argue that this would only be used as a last resort, in a crisis situation, and that, of course, the robot would be controlled by a human officer under the supervision of a senior police commander.
Still: Blowing up bombs to kill people is really not something that many of us want to see as part of local policing.
From the letter, which you can read here:
There is no basis to believe that robots toting explosives might be an exception to police overuse of deadly force. Using robots that are designed to disarm bombs to instead deliver them is a perfect example of this pattern of escalation, and of the militarization of the police force that concerns so many across the city.
SFPD’s proposal would allow officers to send these robots to all arrests and all searches with warrants, and to protests if police decide that ‘exigent circumstances’ or other flexible justifications in the policy apply. Next, it would allow the Chief of Police, or two other top officials, to authorize use of a robot for deadly force if:
“[1] risk of loss of life to members of the public or officers is imminent
and [2] officers cannot subdue the threat after using alternative force options or de- escalation tactics options, **or** conclude that they will not be able to subdue the threat after evaluating alternative force options or de-escalation tactics.”
Thus, given that “or” (emphasis added), this policy allows police to use robotic deadly force without first attempting alternatives.
What really worries me, and hasn’t been much a part of the discussion so far is this:
San Francisco is already allowing cars without drivers to offer taxi rides. At some point, probably in the next five years, someone is going to develop an AI program for police robots to do threat assessment. You know that will happen.
At that point, the cops will ask for that program—of course, only as a pilot, and with human backup at all times, just as the robo-cars were just a pilot with human backup at all times—and pretty soon it will evolve to entirely robotic decisions to use lethal force.
Now we are in the real Robocop arena.
What could go wrong? (Warning: This is violent, gross, and scary. Just like killer robots).
If the supes decide to allow this to go forward, I suspect there will be a measure on the next ballot to overturn it.
