SAN FRANCISCO (AP) — San Francisco supervisors voted Tuesday to put the brakes on a controversial policy that would let police use robots for deadly force.

The Board of Supervisors voted unanimously to explicitly ban the use of robots in such fashion for now. But they sent the issue back to a committee for further discussion and could allow it in limited cases at another time.

It’s a reversal from last week’s vote allowing the use of robots in limited cases. The police said they had no plans to arm the robots with guns but wanted the ability to put explosives on them in extraordinary circumstances.

Last week’s approval generated pushback and criticism about the potential to deploy robots that can kill people. Several supervisors joined dozens of protestors outside City Hall on Monday to urge the board to change course.

Some supervisors said they felt the public did not have enough time to engage in the discussion about whether robots could be used to kill people before the board first voted last week.

The vote was the result of a new state law that requires police departments to inventory military-grade equipment and seek approval for its use.

Supervisor Dean Preston, who voted against the policy last week, said the spirit of the law is to make sure “strong feelings people hold” can be heard by public officials. He argued the board failed to allow enough time for that.

But others said nothing substantive had changed since the board made its vote and the policy should hold.

The policy approved Tuesday would allow police to use robots to check out potentially dangerous scenes so that police can stay back.

“Having robots that have eyes and ears and can remove bombs, which happens from time to time, is something that we want the police department to do while we continue to have this very controversial discussion,” said Supervisor Aaron Peskin, who brought forward last week’s motion around the use of robots.

The new policy needs another vote to take effect.