The spectre of Bill C-27 hung over the Competition Bureau’s AI-focused summit, where international watchdogs warned against repeating Web 2.0 era mistakes.
Matthew Boswell
“We probably need to be making more efforts than we’re making to sync-up the approach around the world.”
Commissioner,
Competition Bureau Canada
The annual event brought together academics, business leaders, and regulators to discuss issues that impact market competition in Canada. The rapid expansion and adoption of AI, as well as the perception that regulators “just don’t get it,” sparked the theme of this year’s conference, Competition Bureau Commissioner Matthew Boswell explained in his opening remarks.
In a tone-setting kickoff for the day, University of Waterloo economics professor Joel Blit provided attendees with an ‘AI 101’ lesson where he predicted that a small number of firms will “dominate” the AI industry within a decade—much like Amazon, Facebook, and Google did in the internet age.
In a pre-recorded keynote address, United States (US) Federal Trade Commission (FTC) Chairperson Lina Khan echoed that concern and warned against allowing only a handful of powerful companies to dominate the market, much like what happened during the early Web 2.0 era. One of the internet giants to emerge from that era, Google, has been facing a slew of anti-trust complaints in recent months, including from the FTC.
In a later panel discussion between international regulators, David Lawerence, a policy director for the antitrust division of the US Department of Justice, said that regulators look back at the “hands-off” philosophy during those dawning days of digital markets with a “twinge of regret.”
Lawrence added that a dominant firm hasn’t yet emerged in the nascent AI era to create an exclusionary ecosystem that threatens the opportunity for other entrepreneurs and innovators, so anyone from any background can build an AI company.
“We’re in that moment, and we want it to remain that way,” Lawrence added. “We need to be reassessing the level of oligopoly we’re comfortable with.”
The moment may not last forever, given the $150 billion in capital expenditures big tech has invested in AI over a one year period, Forbes reported last month. Other international regulators present for the discussion raised similar concerns about potentially anti-competitive practices novel to the AI era.
Carlota Reyners Fontana, the European Commission’s antitrust director, and Ori Schwartz, competition head for the Organisation for Economic Co-operation and Development (OECD), both suggested re-exploring the definition of a merger with the advent of investment agreements that forge strong connections between big tech firms and AI companies. The regulators specifically mentioned the partnerships between Microsoft and OpenAI, as well as Google and Samsung, as examples.
Reyners Fontana also suggested that there could be scrutiny of “acqui-hires.”
“We are concerned that maybe they are designed [to escape merger control], and therefore we look at what core assets might have been transferred,” Fontana said. “[We also look at] whether that transaction would lead to the loss of market presence of the target company.”
Go fast vs. get it right
While the international regulators spoke explicitly about their concerns with the emerging AI market, member agencies of the Canadian Digital Regulators Forum (CDRF) took a wait-and-see approach at a roundtable discussion centered on learning the mechanics of the new digital market.
The CDRF is a cooperation initiative between Boswell’s Competition Bureau and leaders from the Canadian Radio-television and Telecommunications Commission (CRTC), the Privacy Commissioner of Canada, and the recently added Copyright Board of Canada.
Boswell kicked off the discussion noting that the CDRF was created because it’s “a very difficult task” to stay up to speed with the rapid pace of change, and that the proliferation of AI made the work of the member agencies overlap more than ever.
CRTC chairperson Vicky Eatrides said that the overlapping issues and AI’s spike in use provided an opportunity for the leaders to learn from each other.
“You have to work towards where the puck is going and not to where the puck has been,” Eatrides said.
Copyright Board vice-chair and CEO Drew Olsen said he thinks stakeholders and regulators are just now starting to grapple with AI, adding that the copyright board is still trying to understand the changes in market dynamics occurring in the creative sector. Olsen added that he was looking forward to discussions on synthetic media, such as deep fakes, questioning whether regulators have a role to play in “helping people trust what they see on the internet.”
RELATED: AI regulation is coming. What’s Canada’s approach?
The fate of Bill C-27 and its Artificial Intelligence and Data Act (AIDA) provision hung over much of the day’s conversations. The act was described as both “stalled” or “snarled” throughout the day. However, Privacy Commissioner Phillipe Dufresne defended the process, stating that Parliament is “doing its work” by listening to the discussion and debate around the bill. He also noted that there is a balance between legislators moving quickly and a desire to get legislation that will “stand the test of time.”
“There’s lots of proposed amendments, there’s lots of issues, and I hope to see this move forward and I hope to see Parliament give it the priority that it deserves,” Dufresne said. “Time is a precious commodity, it’s always a challenge, especially towards the end of a minority Parliament.”
Dufresne added that organizations that design AI must justify their processes. He went on to explain that AI doesn’t exist in a legal vacuum and that AI developers have current obligations to uphold under existing regulations.
“If [AIDA doesn’t pass] for whatever reason, we are acting, we have ongoing investigations, we have issued statements on AI,” Defresne added. “We’re going to continue to use the tools that we currently have.”
While Boswell didn’t have any prescription for the legislation, he wrapped up the panel by saying the call for regulation from ‘AI godfather’ Geoffrey Hinton, and his warning that the “unchecked expansion of AI is an existential threat to humankind,” caught his attention. Earlier this month, Hinton joined a group of AI scientists to sign yet another letter calling for controls against the malicious use of AI that could “lead to catastrophic outcomes for all of humanity.”
“There are some serious issues and, in my personal opinion, we need to be moving forward on regulation in this country and around the world,” Boswell said.
“We probably need to be making more efforts than we’re making to sync-up the approach around the world, because a lot of the AI is being driven by large tech companies that have more money than many countries, and we need to have some sort of broad approach to it.”
Feature image courtesy Competition Bureau Canada via LinkedIn.