OC Transpo is looking into using artificial intelligence to enhance the effectiveness of its CCTV system on buses as part of the transit company’spublic safety plan.
Such a system would detect assaults, vandalism, loitering and other emergencies in real time, then alert special constables or other responders.
The full application of AI and the public safety strategy is expected to begin in 2026, the City of Ottawa told Capital Current in an email.
Barrhaven East Coun. Wilson Lo, a member of Ottawa’s transit committee, told Capital Current that a committee meeting at city hall on Nov. 24 was the first he had heard about the safety plan in detail.
“There’s a lot of caution we’ll have to take with the implementation — not only with how we use the technology, but who our vendor is,” said Lo.
The main focus of AI is to “comb through footage” and “look for truly unusual behaviour,” said Lo, adding “staff will have to be very careful in developing the software that we get and deciding what they actually want the AI to look for.”
Stittsville Coun. Glen Gower, the transit committee chair, said in an email the new technology is being explored with the hope of identifying possible issues in the transit system and to more effectively deploy special constables. Extensive testing will be done before deployment.
Lo said that if the technology is a success, it could be brought to other public spaces. “If this is successful, transit is the best place to trial this before you could expand into a larger public setting,” he said.
There is public uncertainty about the impact AI will have on everyday life and personal privacy.
According to the Global Public Opinion on Artificial Intelligence survey, conducted by the University of Toronto, violations of citizens’ privacy was among the top concerns about the impacts of AI, right after misuse “for nefarious purposes.”
OC Transpo rider Kyl Manke says of AI, “It’s a slippery slope … so [I don’t feel ]super comfortable about that.” He said he also had concerns regarding data storage.
“I want to know how it’s going to be implemented. I want to know the difference between having an AI tool and having the system right now. What are they trying to move the needle on?” said Lara Shantz, a board member of Ottawa Transit Riders, a transit advocacy group in Ottawa.
Some experts doubt the system’s ability to properly identify threats and report them effectively in real time.
“AI can identify a threat, but can’t always identify the kind of threat. If someone is having a drug overdose or a psychotic episode then they require a critical response team as opposed to police pulling their guns out,” said Blair Maddock-Ferrie, an AI ethics researcher at Carleton University.
One of the main goals of using AI is to “hopefully deter some people from causing incidents,” said Lo. However, Shantz and Maddock-Ferrie say that the financial investment of integrating an AI system may not be the best way to curb problems.
“If we want to talk about a safer system, having a higher ridership, just having more people around, more eyes, more bodies, more bystanders, will actually make more people safer than some AI tools will,” said Shantz.
“I think [the money spent on] AI should be spent more on things that actually work, according to data, like community police and increased efforts for homeless shelters — solutions that address the root cause of the problem,” said Maddock-Ferrie.
