AI and Elections: Governments Must Enact Legislation to Protect Democracies

With Canada’s federal election campaign now in full swing, Catherine Régis, director of innovation at IVADO and professor of law at UdeM, is calling on the government to adopt rules to better regulate the use of AI in the electoral process.

In an open letter published in La Presse, co-written with Florian Martin-Bariteau, professor of law at the University of Ottawa, Ms. Régis reminds us that AI represents a serious threat to democracy in the absence of appropriate safeguards.

“In the past year, it has been used to attack computer systems of Romanian electoral institutions; to publish hyper-realistic fakes to undermine the candidacy of Brazilian female politicians; and to discourage Americans from voting during the Democratic primaries in New Hampshire.”

The letter is co-signed by Sophie Fallaha, executive director of CEIMIA; Christian Gagné, director of the Institute for Intelligence and Data; Lyse Langlois, director of Obvia; Rémi Quirion, chief scientist of Quebec; and Luc Vinet, executive director of IVADO.

This initiative is based on a policy statement published in February to preserve the electoral integrity of democracies. Specifically, the document proposes:

1. Modernizing regulatory frameworks with the adoption of clear rules framing the use of AI during elections.
2. Adoption of codes of conduct for the use of AI by political parties.
3. Establishment of independent teams to monitor electoral integrity and prepare public response plans in the event of AI-fueled threats to elections.
4. Development of an International AI Electoral Trustkeepers and international legal assistance protocols to respond to cases of AI-based electoral interference.

Earlier this year, the two specialists in global AI governance had documented and analyzed, in the report AI and Democracy, attacks against electoral processes that occurred around the world in recent years.