GATINEAU – As part of the unrelenting effort by Canada and other jurisdictions to put an end to fraud and scam calls, last December, the CRTC launched a proceeding that would see the Implementation of universal network-level blocking of calls with blatantly illegitimate caller identification.
The deadline for telcos to have that in place, if they don’t already, is December 19th.
This summer, however, Bell suggested going even further and, as Cartt.ca reported, applied for Commission approval for a pilot project which would use artificial intelligence to build on universal network blocking tools to stop certain verified fraudulent and scam voice calls on a trial basis.
After some back and forth the deadline for interventions was set on October 15, 2019.
As Iristel said, Bell’s request was triggered “because section 36 of the Telecom Act specifically prohibits carriers from interfering with the content of traffic carried by a carrier without permission of the Commission, Bell must seek this permission from the Commission before it can test or implement its call blocking mechanism on commercial voice traffic.”
While most endorse the efforts by Bell, some have strong reservations about calls other than originating, transiting and terminating on Bell’s networks.
For example, “Allstream endorses the goal of reducing fraudulent and scam calls and does not object to Bell blocking such calls it identifies as originating from Bell customers. However, Allstream is concerned and requests extreme caution when allowing a carrier to block third party carrier calls based on a confidential algorithm.”
All mention or decry the fact that the details of Bell’s proposed methods that were filed with the CRTC in confidence and want to know more. Iristel proposes, “Should the Commission decide to allow Bell to block certain traffic, that all carriers whose traffic will be subject to Bell’s filtering technology be afforded the opportunity to thoroughly examine the technical details of Bell’s proposed blocking service.”
The Internet Society Canada Chapter also raised privacy issues: “Given the scope outlined, it is clear Bell will be collecting data from all calls, not just calls from Bell customers, which raises serious privacy concerns. With no opt out mechanism, the calling data for a customer of a CLEC could be included in the Bell ‘AI’ system without any consent or opt-out ability.”
Interestingly enough, Cartt.ca has been told by a Privacy Commissioner’s representative that it has decided not participate at this time.
“Given the scope outlined, it is clear Bell will be collecting data from all calls, not just calls from Bell customers, which raises serious privacy concerns.” – Internet Society of Canada
Telus, for its part, says Bell must “maintain a redress procedure which corrects for any instances of legitimate calls being blocked, whether these instances are identified by Bell’s own customers or by another service provider, including cases where a number used for legitimate calls is also being used by a bad actor for fraudulent calls. Second, Bell must not block a call to 9-1-1 under any circumstances.”
A false-positive would occur when a legitimate non-fraudulent call is accidentally blocked by Bell’s proposed AI-based technology and Bell must implement a procedure to redress the situation. As Telus wrote: “The redress procedure must unblock the number of a legitimate caller without delay, even if fraudulent calls are being made by other parties using that number. In response to Commission staff requests for information regarding false positives, Bell stated that “if in doubt,” it will not block.
Telus supports the fact that details of the methods used to block call should not be made public so not to inform the fraudsters and allow them to bypass the system.
“As always, the challenge is to balance the significant risk of signaling to bad actors our detection mechanisms and thereby providing a 'how to' road map that could be used to thwart this proposed anti-nuisance calling proposal against the need to provide sufficient notice of the proposal to potentially impacted parties,” reads Bell’s original application.
The conclusion belongs to Telus. “Mandated blocking will only stop a small portion of nuisance calls. More sophisticated blocking or filtering mechanisms are required if Canadians are to be protected from telephone fraud. The trial proposed by Bell will explore an additional avenue of fraud prevention. However, no network-level blocking system can be free of false positives and there will be an inherent risk that legitimate, and possibly important, calls will be blocked.
“Rather than holding Bell to an unachievable standard of zero false positives, the Commission must accept that there will be a certain number and consider whether the public benefit of the fraudulent call controls outweighs the potential harm of lost or delayed communications,” reads the Telus submission.
The deadline set last December for companies to implement a universal network-level call blocking system to help cut down on the number of unsolicited and illegitimate calls that Canadians receive still stands and Bell, as well as Quebecor and others, have said they would implement on a timely fashion.
