Listen to this article
Estimated 4 minutes
The audio version of this article is generated by AI-based technology. Mispronunciations can occur. We are working with our partners to continually review and improve the results.
Halton’s police force says it will begin to use an artificial intelligence program in hiring and training 911 call takers.
The police service is using American AI firm GOVWORKX’s technology to help reduce the time it takes to evaluate candidates by 50 per cent and make hiring “more data driven,” according to a news release Friday. The technology will create emergency call simulations to help candidates improve their crisis management and communication skills, it said.
Previously, training and hiring candidates took up to a year, according to its 911 communications manager, Brian Dodd.
Halton police say they receive around 1,400 applications every year, and offer a five-week training course on top of 360 hours on the floor to train call takers. The process repeats when they train to be dispatchers.
If candidates aren’t successful in the end, the force “actually lost a year of time,”
-
Toronto police are turning to AI to eliminate wait times for non-emergency calls
-
Toronto police used Clearview AI facial recognition software in 84 investigations
“That’s where this technology is going to be very efficient,” Dodd said to CBC Toronto. “If we’re hiring the right candidates at the beginning with a high probability of success through training, a year from now, the goal is that they will all still be here and be signed off as 911 communicators.”
The same AI software will help the police train the successful candidates, and continue with “automated quality assurance” through their career
Regulations need to come soon: expert
Dodd said that the police force’s new AI tool will always have a “human in the loop,” but a legal expert says that can create a weak system that needs more guardrails around it.
“Humans working under pressure in a really fast-paced environment and a heavy caseload can oftentimes defer to AI,” said Ryan Fritsch, a lawyer for the Law Commission of Ontario leading law reform projects on AI in Ontario’s criminal justice system.
“They can say, ‘I assume the AI is right’ and just rely on that so they can get to their next call, or investigation, or prosecution.”
WATCH | Use of AI in policing a growing concern for some in Canada:
Use of AI in policing a growing concern for some in Canada
August 26|Duration2:18
Many U.S. law enforcement agencies are using integrated AI technology, including licence plate readers and facial recognition, along with real-time crime centres that pull all that information into a central hub. Some Canadian police departments are also quietly growing their network of AI-powered tools.
While the Employment Standards Act says employers that use AI to screen, assess, or select job applicants need to disclose in their public postings that they’re using AI, Fritsch says Ontario has no specific laws about using AI in the criminal justice system.
“The concern is that the AI can get it wrong,” said Fritsch to CBC Toronto. It can “misinterpret cultural or linguistic cues, make mistakes, mischaracterize events, and get facts wrong too.”
“So when you’re talking about stuff that can end up as evidence in a police investigation or in a court proceeding, that’s really high risk.”
- Video
Canadian police quietly deploying AI-powered cameras, sensors
-
Increasing AI use in Canadian courtrooms carries risk of errors, penalties: lawyers
Fritsch says the province needs AI-specific legislation for using AI in criminal justice, and systems considered high-risk that are using AI need to be subject to mandatory assessments and reporting to police oversight agencies.
“It’s so that everyone knows what’s being used, how it’s being used, and its reliability.”
CBC Toronto reached out to the Minister of Attorney General for comment.
Further AI integration plans
Halton police’s announcement comes a year after the force started using another technology, called SARA AI, to answer specific types of non-emergency phone calls.
SARA AI reduced 911 wait times by almost two seconds a call, according to the police news release.
“On a 911 call, two seconds is very important,” Dodd said.
Halton police plan to introduce a second phase for SARA AI in January
SARA AI handled over 95,000 calls this year, and spent over 1,000 hours talking to residents, according to police.
