Canada joins global partnership to promote ‘responsible’ AI
Fourteen nations collaborating to promote human rights, diversity, inclusion and innovation
WATERLOO REGION — Canada has joined an international partnership of more than a dozen nations to promote the ethical and “responsible” advancement of artificial intelligence. Announced on Monday, the Global Partnership on Artificial Intelligence (GPAI) will look to apply shared values such as human rights, diversity, inclusion, innovation and economic growth to the technology.
Local experts agree this first-of-its-kind partnership is an important step toward establishing an ethical framework around the future development of AI, but it remains to be seen just how effective the partnership will be at curbing its malicious or destructive potential.
“I think it’s necessary and a good start, but how will it be implemented and what influence could they have, especially on private firms?” questions Branka Marijan, senior researcher with Waterloo-based Project Ploughshares, a nonprofit organization dedicated to global disarmament and a member of the international Campaign to Stop Killer Robots.
“It’s important that countries collaborate on this. One country can’t do it alone,” added Peter van Beek, a co-director of the Waterloo Institute for Artificial Intelligence.
AI allows computers to learn from experience by processing huge amounts of data to find hidden relationships, patterns and anomalies so that over time they can make better predictions and learn how to perform tasks without human instruction.
The opportunities for health care, entertainment, science and a slew of other industries is profound, but the proliferation of AI also raises important ethical questions — most notably its use by military or police forces around the world to develop massive surveillance networks or even “smart” weapons systems such as tanks or missiles.
GPAI members include Australia, Canada, France, Germany, India, Italy, Japan, Mexico, New Zealand, the Republic of Korea, Singapore, Slovenia, the United Kingdom, and the United States of America. The European Union is also in negotiations to enter the partnership.
Their mandate will be to facilitate international collaboration by bringing together experts from industry, civilians, governments and academia to discuss four main themes: Responsible AI, Data Governance, Future of Work, and Innovation and Commercialization.
GPAI will also investigate how AI can be leveraged to recover from COVID-19.
Two research centres will be established to help carry out this work — one in Canada at the International Centre of Expertise in Montréal for the Advancement of Artificial Intelligence, and a second centre in Paris, France.
The two countries have been two of the main drivers behind the formation of the group, beginning in June 2018.
Marijan noted how one nation is notably absent from the list of participating countries.
“If the aim is to build a normative framework for AI globally, you need some way to engage with China,” she said.
Van Beek agreed. “China has gone all-in on AI,” he said. They’ve developed their own guidelines for managing the development of the technology, “but many are skeptical,” van Beek added, noting how AI technology such as powerful facial recognition software is now part of every day life in China.
Marijan would also like to see private companies currently at the forefront of AI development — Google and Amazon, for example — included on the list of members.
“They need to be engaged somehow, because they have a lot of power.”
Yet van Beek argued against their inclusion, saying their presence could taint how the group is perceived and they may have a negative influence over the decisions or regulations being developed.
“I think it’s good they’re not at the table. They can’t self-regulate and don’t do it, so it needs to be imposed on them,” he said.
James Jackson is a Waterloo Region-based reporter focusing on business and technology for the Record. Reach him via email: jjackson@therecord.com