Skip to main content

BCE Inc. is proposing a new call-blocking system that would use artificial intelligence to screen for potential phone scammers, though the plan is raising questions about how it would work and how the resulting data would be used.

The Montreal-based telecom giant plans to launch a 90-day trial to block calls from recognized scam phone numbers – the first system of its kind in Canada – according to an application posted to the Canadian Radio-television and Telecommunications Commission in late July. BCE currently allows customers to avoid unwanted calls by adding phone numbers to a block list or by blocking all calls except those from recognized contacts.

The new system would block known fraudulent numbers across the network and incorporate AI and machine learning to identify suspicious activity.

“We want to take these measures to increase confidence in the voice system because if ordinary Canadians are reluctant to pick up their phones because they don’t trust that the next call may be spoofing them or actively trying to defraud them, it’s bad for Canadians and it’s bad for business,” said Jonathan Blakey, BCE’s assistant general counsel of regulatory affairs.

The increase in fraud calls – such as the scammers identifying themselves as the Canada Revenue Agency, a ruse that last year the RCMP said garnered 4,000 victims who lost $15-million – has attracted the attention of regulators. In December, the CRTC directed telecommunications companies to introduce network-level call-blocking systems by the end of this year that rejects calls with caller-ID information that exceeds 15 digits or does not fit the standard 10-digit phone number.

The U.S. Federal Communications Commission announced in June that it would allow carriers to implement network-level call blocking, a stark shift from its previous position on the issue, according to PC Mag’s lead mobile analyst Sascha Segan.

“Carriers were not permitted to do network-level call blocking because it was considered an interference with the freedom of the callers,” he said. “A carrier blocking calls en masse was considered a carrier almost taking an editorial policy to its calls, as opposed to now when it’s considered just managing the networks so it’s not all spam.”

But Fenwick McKelvey, a Concordia University communications professor, said BCE did not provide enough public information on the AI proposal, which is posted on the CRTC website. BCE redacted details about how the blocking system works, citing concerns with revealing information that could allow fraudsters to figure out how to circumvent the tool.

Prof. McKelvey submitted a request on Aug. 8 asking that the CRTC allow the public to ask for additional information, adding that the AI component calls into question how the system decides who is blocked, how the data is used and whether the system can handle privacy risks.

“This is the start of what I think will be a new phase of regulatory debate on how we use AI effectively to manage telecommunications networks. It’s an important one to have and hopefully not one lost as it is now, buried in a technical submission to the CRTC," Prof. McKelvey said.

The CRTC said that it did not recognize the issue as a matter of public interest, but that the Commission would address the concerns when assessing the application, according to a notice sent to Prof. McKelvey on Monday.

Your time is valuable. Have the Top Business Headlines newsletter conveniently delivered to your inbox in the morning or evening. Sign up today.