Don’t rush new laws for AI, data conference hears

TORONTO — There are ways to regulate artificial intelligence and the amounts of data that it produces, but introducing unnecessary new rules to protect privacy when current laws should suffice is not one of them, the Big Data Toronto conference heard Tuesday. Trying to rush in new laws to manage new technologies that regulators may not understand, such as AI, will hamper companies trying to deliver those technologies and may lead those laws to inevitably lose relevance in the future, panelists at the Big Data Toronto conference argued. “How do we take what's in place today and apply it and be creative there, because otherwise you do potentially introduce new things that might make sense today, but the moment they're passed, that law is already out of date,” said panelist Suzanne Morin, enterprise chief privacy officer at Sun Life Financial. “So it's not about not regulating and let's see where things go, but rather it's apply the existing rules that exist already and try not to be technology specific,” she said. Morin was one of three panelists tasked with providing insights on regulation in AI, including privacy concerns related to data collection, and the role government can play at the Big Data Toronto 2018 conference. Morin emphasized the quality of Canadian privacy laws and how far ahead they are — being principles-based — compared to the more prescriptive European Union's new General Data Protection Regulation (GDPR). The GDPR sets out specific rules for user consent and fines for breaking its data protection laws, which amount to four per cent of global revenues. Carole Piovesan, a lawyer at McCarthy Tetrault LLP, said that kind of blackletter law — rules that are not easily malleable — can actually deter new technologies for fear that they could run contrary to the law. “We have to be careful with blackletter law because it doesn't change easily,” Piovesan said. “You do it once and you try not to do it for a long time. But there is room for that as we get closer to a more sophisticated, more mature technology and industry.” For that, Piovesan did argue on the side of the right kind of regulation that will help give those industries the concrete guidance they need. “Sometimes regulation makes your life more difficult, but sometimes it creates certainty in an industry and sector, which allows the sector to mature,” Piovesan said. “So smart regulation that is slow to grow but there to protect is often a good draw to create a more mature sector and a more mature industry.” Lee Brotherson, director of security at financial services company WealthSimple, rounded out the panel. He also said while regulation can help guide industries like his when it comes to certain things, like what to use and not use for identification purposes, he also said the regulatory approach should provide a general guide and not be too narrow in focus. “If they can take things to a high level, that you need to meet the spirit of this principle, it will allow innovation to move forward, which I think is very much the area AI is interested in,” Brotherson said. But the message that underpinned the conversation and resonated through the hour-long conversation was the need to ensure that adequate steps are taken to ensure not only privacy, but transparency and accountability as well. Morin noted that industries need to be moving beyond the privacy policy and toward “transparency 2.0,” which includes proactively giving clients all the information they need to understand those privacy policies. Piovesan said while the privacy by design is a strong jumping off point, she laid out her own version of it, called “transparency by design.” The idea is that all things related to data collection and its use be made not only understandable, but that all information is given to the user so they can make informed decisions on for what they’re giving up their information, such as apps that they use. “Sometimes that information involves that your data will be used in a secondary manner, and here are the parameters in which they will likely be used,” Piovesan said, adding sometimes that would require going back to the customer again to get consent. Keeping with the spirit of the principles-based approach, that level of transparency would encompass all sectors and industries, she said. “Really think about transparency and interpretability from the outset as you start to collect that data,” she suggested. “Think about how you're going to use that data, then use that data, and then reuse that data...that's critical to not only being compliant with the law and your obligations under privacy law — it's also critical to maintaining that trust relationship with customers and clients.” — With reporting by Ahmad Hathout at ahathout@thewirereport.ca and editing by Anja Karadeglija at akarad@thewirereport.ca

Don’t rush new laws for AI, data conference hears

News | June 12, 2018

Unlock all the Canadian telecom, broadcasting and digital media news you need.

FREE TRIAL

SUBSCRIBE

Inside The Wire Report

Information

Hill Times Publishing

Get in Touch With Us

Don’t rush new laws for AI, data conference hears

News | June 12, 2018

Unlock all the Canadian telecom, broadcasting and digital media news you need.

FREE TRIAL

SUBSCRIBE

Related Stories

Four cyber bills dead on the order paper as Parliament prorogued

Feds launch AI compute strategy

Canadian media companies sue OpenAI for copyright infringement, could cost $322 billion

OPINION: Government agencies using AI must implement governance programs, or risk losing trust

Reuse & Permissions

Inside The Wire Report

Information

Hill Times Publishing

Get in Touch With Us