An international panel established to set ethical guidelines for the use of artificial intelligence (AI) finally launched on Monday, nearly two years after France and Canada announced plans to form the global group.

The Global Partnership on Artificial Intelligence, will bring together experts from the private sector, academia and non-profit organisations to conduct research and pilot projects on AI.

The panel’s objective, as set out by founding members Australia, Canada, the EU, France, Germany, India, Italy, Japan, Korea, Mexico, New Zealand, Singapore, Slovenia, the UK and the US, is to “bridge the gap between theory and practice on AI policy”, according to a statement.

The group’s star recruit is deep learning pioneer Yoshua Bengio, founder and scientific director of the Mila research institute in artificial intelligence in Quebec, who will co-chair a working group on responsible AI. It is hoped the presence of Bengio, one of the so-called AI “godfathers” who helped propel Canada into AI, will lend the panel immediate credibility.

Canada’s president Justin Trudeau and French president Emmanuel Macron announced plans for this new international body back in 2018. The idea was to model it on the UN’s Intergovernmental Panel on Climate Change, which helped establish consensus on the climate crisis and recommend possible responses.

In pushing for a standing forum for AI, the two leaders were responding to mounting concerns that AI, while potentially beneficial, could also lead to dystopian scenarios of privacy invasion, discrimination or even warfare.

The international panel finally began to take shape last month when the Trump Administrationin the US dropped its opposition to joining.

The US, though concerned early on that the group would hamper AI development by setting rigid standards or regulations, is increasingly wary of seeing a world shaped by China’s view of the technology.

On announcing America’s sudden support for the panel, the White House's chief technology officer, Michael Kratsios, suggested the US wants to help guide AI to serve democratic interests, rather than authoritarian ones.

“The pandemic has made clear why AI development aligned with privacy, freedom, and civil liberties is such an imperative. It is critical that America stand alongside those who share and promote our values,” said in a tweet.

Hubs in Montréal and Paris

Coordination of GPAI will come mainly from researchers based in Montréal and Paris.

The Canadian government announced the official opening of the International Centre of Expertise in Montréal for the Advancement of Artificial Intelligence, which will host the first annual group plenary in December 2020. The government has committed to investing up to C$15 million over five years in the group; the contribution of the other 14 member countries is still to be revealed.

France announced a corresponding hub in Paris based in the National Institute for Research in Computer Science and Automation.

Both centres will work with the group’s secretariat, run by the Organisation for Economic Co-operation and Development. The OECD has established a network of AI experts to advise members on policy, and it has produced a set of AI principles endorsed by more than 40 countries.

The global panel will draw up guidelines on four broad themes, including responsible AI, data governance, the future of work, and innovation and commercialisation. A full list of experts to work on these topics is being finalised.

The Montréal centre will focus on responsible AI and data governance. The centre in Paris will address the future of work and innovation and commercialisation.

Researchers will also investigate how AI can be leveraged in the immediate term to respond to and recover from COVID-19.

The working groups, in turn, are to be overseen by a series of three committees: a ruling council that includes government ministers, a steering committee, and a multi-stakeholders experts’ group plenary that includes public and private experts.