Cops Are Asking to Kill People With Robots. What Could Go Wrong?
Multiple major cities including San Francisco and Oakland this year have considered obtaining armed police robots that can kill people.
This piece is a commentary, part of The Appeal’s collection of opinion and analysis.
The U.S. military has been killing people with robots for decades now, and the nation’s local police now seem eager to get in on the action.
Drone strikes abroad have become so commonplace that the mainstream news media barely bothers to cover them anymore. For years, the military has also been using bomb disposal robots, which are basically glorified radio-controlled cars with tank-treads that can be used to safely dismantle explosives from a distance. These robots have proliferated across local law enforcement in recent years, as military technology so often does. In 2016, the Dallas Police Department took the extraordinary step of mounting a bomb to its bomb disposing robot and using it to kill a mass shooting suspect.
Since that incident, civil liberties groups have raised concerns that law enforcement would seek to use robots to kill people in greater numbers. Police are now showing that those fears were not unfounded. In the last few months, at least two different cities have officially proposed allowing police officers to kill people using robots. In September, the Oakland Police Department discussed the possibility of using a Remotec Andros Mark V-A1 robot equipped with a shotgun to kill suspects during “high-risk, high-threat” events, as local journalist Jaime Omar Yassin and The Intercept reported. After public outcry over potentially being shot to death by a ‘roided-out Mars Rover, the department abandoned the idea. For now anyway.
The same debate is now playing out across the Bay. On Tuesday, the San Francisco Board of Supervisors voted to enact a policy that will allow its officers to kill people with robots. The ordinance’s language states that the robots “will only be used as a deadly force option when risk of loss of life to members of the public or officers is imminent and outweighs any other force option available to SFPD.” The department’s policy had previously explicitly banned police from unleashing drones and robots that could use force against people.
After one San Francisco supervisor, Dean Preston, issued some mild pushback to the proposal, the San Francisco Police Officers’ Association (POA), the city’s police union, posted a scathing Twitter thread claiming Preston did not want to stop mass shooters. The final vote was 8-3.
Other departments seem likely to follow suit. Local cops now own thousands of Andros Mark V-A1 bots, according to The Intercept’s report last month. Local police have also increasingly used drones to monitor all sorts of activities, from peaceful protests to drug deals.
The proliferation of police robots has sparked pushback from civilians, elected officials, and civil rights groups alike. In 2021, residents and local politicians castigated the New York Police Department for its use of a so-called “Digidog”—an unarmed, four-legged, robotic “dog” built by Boston Dynamics—after footage of its deployment in two separate incidents went viral. The city has since canceled a $94,000 contract to lease the robo-dog. While Boston Dynamics was one of multiple companies that signed an open letter last week condemning the use of armed robots, other companies haven’t drawn such a line in the sand. As TechCrunch noted this week, a Philadelphia company named Ghost Robotics sells its products to the U.S. military and seems totally fine with the strapping of rifles to its robo-dogs.
On Monday, the Electronic Frontier Foundation (EFF), a digital privacy nonprofit, issued a scathing statement slamming the San Francisco Police Department’s request for killer robots.
“This is a spectacularly dangerous idea and EFF’s stance is clear: police should not arm robots,” the organization said. “Police technology goes through mission creep–meaning equipment reserved only for specific or extreme circumstances ends up being used in increasingly everyday or casual ways. We’ve already seen this with military-grade predator drones flying over protests, and police buzzing by the window of an activist’s home with drones.”
As EFF noted, the language governing police use-of-force policies tends to be extremely broad, including for the use of robots. While departments often claim that military technology will only be used in rare circumstances, the actual rules are often written to give cops leeway to do virtually whatever they want with the technology. In San Francisco, the policy would allow the department to deploy armed robots during nearly any situation—cops could use these weapons to kill people remotely so long as they claim to fear for their lives.
Likewise, the American Civil Liberties Union has long warned against the arming of police robots. In addition to the many other reasons these machines could be harmful, the ACLU has noted that officers likely do not have the same tactical or situational awareness when using a remote-controlled device. The simple misreading of a person’s movements or other cues could lead to someone getting hurt—or worse.
“When officers are acting remotely, they don’t have live, 360-degree, multi-sensory, intuitive situational awareness, and their perception of a situation is more likely to be flawed or confused, and force used inappropriately and/or on the wrong targets,” the ACLU warned in 2016 after the Dallas police killing-by-robot. “Signals may also be degraded due to communications and control problems with the robot.”
To get a true sense of the terrifying possibilities of this dystopian future, one need only to look at the military’s propensity for using drones to kill innocent people. Remote weapons make it easier for soldiers to kill people from afar, and news reports have shown us how severe the consequences can be when state actors misread critical evidence.
In 2021, the New York Times released a series of investigative reports detailing horrors committed by military drone pilots. The Times noted that people were often attacked or killed remotely based on misunderstandings or flimsy evidence. In one incident, officials authorized a drone strike on the home of Basim Razzo, an innocent man, after just 95 minutes of surveillance, killing his wife, daughter, brother, and nephew. In those few minutes of monitoring, the Times reported that the military made stunning errors, such as interpreting the opening of a gate or the absence of women as “ISIS activity.” Razzo survived but needed major surgery to correct multiple shattered bones and remove pieces of shrapnel from his body.
Despite these horrors, the U.S. military continues to use remote weapons with troubling frequency. And now, after Americans failed to stop the widespread use of this technology overseas, it may be coming home.
This story has been updated to reflect the San Francisco Board of Supervisors’ approval of the new use of force policy regarding police robots.