The San Francisco Board of Supervisors voted down a controversial policy that would have allowed police to deploy robots capable of using lethal force in extraordinary circumstances, only a week after approving it.

The policy, initially approved during the board's November 29 meeting, would have allowed officers to use ground-based robots to kill when "risk of loss of life to members of the public or officers is imminent and officers cannot subdue the threat after using alternative force options or de-escalation tactics,” according to CNN.

The proposal was part of a broader piece of legislation authorizing San Francisco Police to obtain and use military gear. The vote sends the legislation back to committee, and the board intends to continue debating the use of robots going forward, according to KRON. As a result of the December 6 vote, the language authorizing the robots to use lethal force will be changed to suggest that the robots will not be allowed to use deadly force.

San Francisco police already have 17 robots, Police Magazine reported. A police department spokesperson told Government Fleet they are maintained by its Special Operations Bureau. City Supervisor Rafael Mandelman said in a tweet that the department has been using them since 2010, and has never used them previously to deliver lethal force. His tweet went on to mention that there is only one instance on record of a law enforcement robot being used to kill someone: in 2016 when Dallas police used a robot to kill a sniper after a 45-minute gun battle and two hours of negotiations after the suspect had shot five officers.

The mayor still has to approve the general ordinance before it goes into effect.

About the Legislation

Last week, the board voted eight-to-three to allow the use of the remote-controlled armed robots in extreme situations. According to the meeting minutes, the vote approving the policy came after the language of the proposal was amended during last week's meeting to include that only the police chief, assistant chief of operations, or deputy chief of special operations may "authorize the use of robots as a deadly force weapon."

In an interview last week with CNN, Police Chief Bill Scott said the robots would only be allowed to be operated by officers with specialized training. Explosive charges could be added to the robots to breach fortified structures, or the robots could be deployed to “contact, incapacitate, or disorient” a dangerous suspect without risking the life of an officer, Scott told CNN.

Police said they had no plans to arm the robots with guns, but wanted the ability to put explosives on them in extraordinary circumstances, KRON reported. The approval sparked backlash, with the board collecting responses from the community in large condemning the policy, and a protest outside City Hall on December 5.

The use of robots capable of causing injuries has caused backlash in recent months. In June, Axon halted the development of a TASER-armed drone that could be used for active shooter response after several members of its AI Ethics Board resigned.