San Francisco’s Board of Supervisors voted Tuesday to bar police from using robots for deadly force during extreme incidents. The decision was a striking reversal after the Board gave law enforcement that right on Nov. 29.

The board had received widespread criticism after voting approve a police proposal authorizing law enforcement to deploy remote-controlled, ground-based robots to use deadly force when there is “imminent” risk to life and alternative measures to subdue the threat do not work, the Washington Post reports.

Officials on Tuesday also sent the issue back to a committee for additional review, leaving the policy open to future amendment.

Pressure from activists caused board members such as Gordon Mar (D) to publicly switch their positions ahead of Tuesday’s vote. Mar said Tuesday that he had grown “increasingly uncomfortable” with the precedent the policy would set for other cities and had decided to vote against it.

The debate over whether police can deploy killer robots first emerged in the United States after an incident in Dallas in 2016. After a lone gunman killed five officers during a sniper attack and refused to surrender after lengthy negotiations, police placed explosives on a robot and detonated them to kill the shooter.