Toronto Approves Gunshot-Detecting Surveillance Tech Days After Mass Shooting

Rattled by a string of gun violence this year, Toronto just approved ShotSpotter, a US-based audio surveillance system that purports to detect the sound of gunshots.

|
Jul 25 2018, 2:56pm

On Tuesday night Toronto city council voted to approve new and controversial audio surveillance technology for police that will cover five square miles of Canada’s largest city.

The vote was taken just days after a lone gunman shot and killed two people—an 18-year-old woman and a 10-year-old girl—and injured 13 others in a Toronto neighbourhood on Sunday night in an act of mass gun violence.

ShotSpotter is a proprietary, US-based microphone surveillance system that purports to detect the sound of gunshots by capturing audio snippets, triangulating the source of the sound, and alerting police within 45 seconds. The system is widely used by police in the US, but reports of its effectiveness have been uneven.

Read More: Cities Are Using Hidden Webs of Acoustic Sensors to Detect Gunfire

Community members and civil liberties advocates worry that ShotSpotter’s microphones will impinge on people’s privacy and unfairly surveil racialized communities in Toronto, where more than half of the population identify as visible minorities.

Several city councilors denounced the rushed deliberation process on ShotSpotter—Toronto Mayor John Tory introduced the idea in a surprise motion last week and without public consultation.

In recent months, several broad daylight shootings in Toronto have sparked an intense debate about how to curb gun violence in the city even though the hard data doesn’t show a huge uptick—Toronto has experienced 228 shootings in 2018 so far, up from 205 this time last year. Toronto is frequently ranked as being one of the safest cities in the world.

In response, city staff put forward a $44 million CAD proposal spearheaded by Tory and Toronto police Chief Mark Saunders that includes $30.3 million for community initiatives and roughly $13.5 million for enforcement and surveillance, with $4 million carved out for ShotSpotter. The plan also includes more than doubling the number of CCTV surveillance cameras in Toronto, from 34 to 74.

Ahead of the vote, a group of prominent Black academics, artists, and activists released an open letter to Tory and Ontario Premier Doug Ford calling for the $4 million allocated to ShotSpotter to be spent on community programs instead.

“These [police enforcement and surveillance] initiatives will only further barricade and quarantine Black communities that are already under economic pressure from government policies ushering in an unprecedented era of unaffordable housing and meagre job prospects for young people,” the letter stated.

The Canadian Civil Liberties Association (CCLA) had urged Tory to delay the vote on the policing plan until a legal risk analysis could be conducted.

For the CCLA, the inclusion of microphone surveillance in Toronto’s anti-gun violence plan adds an Orwellian dimension to existing concerns that police resources will be deployed disproportionately to minority neighbourhoods.

Michael Bryant of the CCLA likened this outcome to “carding by audio technology,” referring to the (now restricted) stop-and-frisk-like program of “carding” that effectively amounted to racial profiling in Toronto.

Unless the city places ShotSpotter sensors in wealthy neighbourhoods as well as low-income and racialized areas, “it’s going to create the perception, if not the reality, that particular neighbourhoods are being targeted because of their ethnicity,” Bryant said over the phone.

At Tuesday’s meeting, Saunders did not say exactly how much audio information is recorded by ShotSpotter before and after a gunshot or gunshot-like sound.“I’m sure there is a buffer where it goes back a second or two,” he said. “I don’t think it will grab a lot of information, but the bigger question from a public perspective is that it’s grabbing—it has the opportunity to grab—and I think that’s the biggest concern.”

The Toronto police are not interested in people’s conversations, he insisted.

Still, it’s a fact that voice recordings captured by ShotSpotter’s system have been used as evidence in several court cases in the US, according to the ACLU, where the devices captured “relevant voices” after gunshots.

“That is troubling because six seconds is both a long time and too short to provide any context,” said Bryant. “You can see how it could be easily misused as evidence.”

There are significant questions left unanswered regarding ShotSpotter, despite council’s approval. For one: does it even work?

A 2016 Forbes investigation claimed that 30 to 70 percent of the time, officers responding to ShotSpotter notifications couldn’t find any evidence of gunshots. Police in Oakland, California say the system is expensive and redundant and are aiming to eliminate ShotSpotter. In 2012, the UK city of Birmingham pulled the plug on ShotSpotter after only two out of 1,618 alerts were confirmed as gunshots.

Saunders admitted there will be false positives but noted that “what’s most important is how my officers respond.”

When asked in the meeting if Toronto police had conducted an independent test of ShotSpotter’s effectiveness, Saunders insisted he would not “bring a defective product” to Toronto and had been assured of its reliability.

That and other questions, such as where the audio data from Toronto’s streets will be stored—if it resides on US servers, it can be accessed by US security agencies—and how the program will be overseen will be addressed at a meeting of the Toronto Police Services Board in September. There will be an opportunity for public consultation then, and a privacy impact assessment for ShotSpotter will be completed before its implementation.

But it’s all arguably too little, too late. “Let’s hope that it’s a possibility that the final decision about whether or where it gets put up has not been made,” Bryant said. “But my experience is when a municipal government approves something, they like to spend it.”