On a warm afternoon this past April, yellow school buses rolled toward the Lowertown Community House to deliver children to its after-school program. Moments later, the calm was shattered by reports of an active shooter in the neighbourhood.
The community centre, where kids had been arguing over markers while tired parents signed them out, quickly became an improvised emergency shelter. Staff rushed dozens of confused children inside and locked the doors as police scrambled to locate the shooter.
Minutes later, officers requested access to the centre’s surveillance footage. The resident complied. Today, footage such as this feeds into Ottawa Police Service technology systems, including body cameras rolling out citywide starting in late 2025.
For Mélissa Thibeault-Canas, the social worker on duty, the incident crystallized a growing unease. She was shaken by the shooting and deeply concerned for the children. But the introduction of body cameras and other AI-assisted tools added another worry, giving her the sense that the neighbourhood was becoming a heavily monitored space.
“I understand the need for safety, but I worry that AI in policing could make Lowertown feel even more surveilled,” Thibeault-Canas told Capital Current. She fears an imperfect algorithm might wrongly identify residents and deepen the mistrust that exists in Lowertown, the ByWard Market, Sandy Hill and Vanier.
I understand the need for safety, but I worry that AI in policing could make Lowertown feel even more surveilled.
Mélissa Thibeault-Canas, social worker
As Ottawa police increasingly turn to tools such as facial recognition technology systems, automated video analysis and body cameras, the force promises faster investigations and quicker suspect identification. In Lowertown, where about 40 per cent of residents live below the poverty line, strained police relations and historical over-policing already shape daily life.
The rollout of these technologies is raising urgent concerns about privacy, fairness and civil liberties. Even small AI errors could disproportionately erode trust in marginalized communities.
Experts warn that, without proper regulation, AI tools may do more harm than good.
Sharon Polsky, president of the Privacy and Access Council of Canada, says Canada is unprepared for police use of AI.
“Artificial intelligence is neither artificial nor intelligent — it does not think,” Polsky said. “It repeats patterns from the data it’s trained on. And if that data reflects systemic bias, the output reflects that too.”
Polsky points to the failed Bill C-27, a federal attempt to modernize privacy laws by creating rules for AI systems, including technologies used in policing. The bill died when Parliament was prorogued in January.
As a result Canada remains without privacy protections or a regulatory framework for police use of AI. Experts say this widens the accountability vacuum.
“The bill failed because it had more gaps than safeguards,” Polsky said. “There were exemptions for government institutions, weak oversight and no clear mechanisms for people to challenge decisions made by flawed technologies. That’s a dangerous combination when you’re talking about policing,” she added.
She warns that, without clear laws and oversight, Canadians risk exposure to AI-driven surveillance, wrongful identification and ambiguous handling of personal data.
“We rely on private companies for systems that touch every part of our lives, yet there is almost no accountability, no public disclosure and no way for Canadians to challenge the decisions these tools make,” she said.
Her warning comes as the City of Ottawa is increasing its investment in policing technology.
The concerns of Polsky are also echoed by other civil liberties advocates, who say AI increases the risks of over-policing marginalized communities and could impact the way they’re monitored and treated. They also say that AI could deepen the risk of discriminatory policing.
Algorithms trained on biased historical data put people from marginalized communities at greater risk.
Tamir Israel, Canadian Civil Liberties Association
The Canadian Civil Liberties Association (CCLA) has repeatedly raised concerns about the use of facial recognition and predictive policing tools across Canada.
Tamir Israel, director of the Canadian Civil Liberties Association’s Privacy, Surveillance & Technology Program, says these tools can encode and reproduce decades of discriminatory policing practices.
“Algorithms trained on biased historical data put people from marginalized communities at greater risk,” Israel said. “All this information reflecting historical discrimination is packaged into a so-called neutral algorithm.”
With police services across Canada now experimenting with these tools, biased systems could affect communities nationwide.
The CCLA has described facial recognition as an “indiscriminate, disproportionate and warrantless search” of the public. The organization recommends a moratorium on the technology until Canada enacts independent oversight and transparency requirements, as well as creating avenues for the public to challenge algorithmic decisions.
A 2022 report from the U.S. National Institute of Standards and Technology (NIST) found that some facial-recognition systems misidentified Black and Indigenous individuals 10 to 100 times more often than white individuals. A 2025 study showed that blurry, low-resolution or poorly lit images caused false-positive rates to spike, disproportionately affecting Black individuals, particularly Black women.
These findings highlight the risk racialized residents face of being wrongly flagged or pulled into police investigations.
“These systems don’t just fail in a vacuum,” Israel said. “They fail in ways that reflect who has historically been over-policed.”
The 30 Axon body cameras being deployed by the Ottawa Police Services will record interactions with the public, creating a permanent record for investigations and potential accountability.
The pilot will run into early 2026, with a full rollout planned for 2026–2027, pending budget approval.
Israel cautions that cameras do not automatically guarantee accountability.
“Body-worn cameras can promote transparency,” he said. “But when combined with AI that is able to analyse behaviour or identify individuals in real time, the privacy risks multiply sharply, especially for marginalized communities.”
In contrast, police officials say these tools can be useful in investigating crimes. Supporters say they can speed up responses and help police interpret large amounts of data.
These technologies allow us to identify suspects faster, locate missing persons, and respond more effectively during emergencies. They do not replace human judgment; they support it.
Jean-Claude Lemonde, chief information officer, Ottawa Police Service
Jean-Claude Lemonde, chief information officer for the Ottawa Police Service, said AI-assisted tools help investigators process evidence more efficiently, especially in major cases involving hours of video footage.
“These technologies allow us to identify suspects faster, locate missing persons, and respond more effectively during emergencies,” Lemonde said. “They do not replace human judgment; they support it.”
Lemonde emphasised that any AI-related tools used by the Ottawa Police Service follow provincial and federal guidelines.
“We take privacy seriously and ensure all tools comply with legal requirements,” he said.
Ottawa’s 2026 draft budget increases the OPS budget $484 million with a net operating total of $414.9 million, up $26.1 million over 2025, a 5.0 per cent rise, the force says. The money will be used to hire 25 new full-time positions in 2026 and to fund the full roll out of body cameras.
Police describe the increase as part of its effort to modernize policing in Ottawa “to boost community relationships, visibility, and trust.”
Community advocate Robin Browne of the 613-819 Black Hub says “none of these reforms, including AI, are meant to reduce police violence,” Browne said. “They’re meant to legitimize policing, increase the budget and allow surveillance of vulnerable communities.”
Browne says body cameras and facial recognition software represent a new iteration of this surveillance. Every reform, including AI, he says, comes with a price tag, which is often used to justify further budget expansions.
The police always need more money for whatever the reform is, whether it’s hiring more Black officers, rolling out a district model or implementing AI. … Every reform does the same thing: legitimizes the budget and expands their surveillance capacity.
Robin Browne of 613-819 Black Hub
“The police always need more money for whatever the reform is, whether it’s hiring more Black officers, rolling out a district model or implementing AI,” he said. “AI would cost a whole bunch of money and every reform does the same thing: legitimizes the budget and expands their surveillance capacity.”
“If they really want to build trust, they need to shift significant funding from the police to things that actually keep us safer—things like housing, employment programs, and mental health services. That’s what truly makes us safer,” he added.
This financial shift adds a deeper layer to the debate over transparency and accountability.
The debates aren’t just abstract for community members like Mélissa Thibeault-Canas.
By allocating more money to policing instead of vital community support, Thibeault-Canas says, the city is signalling its priorities are misaligned with community needs.
“We need investment in people, not just cameras and algorithms,” Thibeault-Canas said.
