With Bill C-27, Canada has a golden opportunity to become an international role model in artificial intelligence (AI) leadership. But the slowness of its adoption means that “the window for action is closing quickly”, denounce the 75 signatories of an open letter published this Wednesday in La Presse.
The group is made up of artificial intelligence researchers and academics like Yoshua Bengio, Catherine Régis and Graham Taylor, CEOs of tech companies like Sam Ramadori (Brainbox) and Frantz Santellemy (LeddarTech), and heads of related organizations to AI like Julien Billot, of Scale AI. It is a logical follow-up to the call made by some 20,000 experts on March 29 for a global six-month moratorium on the development of AI. This call also challenged political decision-makers to establish AI governance.
“There is no doubt that AI offers great potential for economic growth and societal opportunities, including to address some of the most pressing health and environmental challenges of our time,” said a statement. straight away in the letter. While these developments are impressive, it is important to recognize that AI systems pose risks that need to be taken seriously and addressed quickly. »
Bill C-27, introduced in June 2022 by the Trudeau government and which is the subject of a standoff with the Conservative opposition which is delaying its passage, is an “excellent” measure in this direction, said in an interview Yoshua Bengio, who contributed to the drafting of the open letter. “We would surely be ahead of other countries. There are things that could be improved […], but there is an urgent need to act, because there is fairly rapid progress in AI. »
The researcher wants Bill C-27 and its third part called “Artificial Intelligence and Data Act”, which aims in particular to prevent the reckless use of this technology, to be passed before the end of the parliamentary session , in mid-June. A year after its introduction, the bill is only at its second reading.
Mr. Bengio recalls that a broad consultation will then be carried out before establishing the precise wording of this law. “You have to start the process right away, which will take time. What I asked the minister is to step up the pace. The two-year consultation process was the initial plan, before ChatGPT arrived. He suggests that certain aspects could be framed by regulations, after the adoption of the law. “Technology moves fast, while it’s hard to change laws. »
The adoption of an AI framework would have economic advantages, underline the signatories of the letter. “If Canada were to become one of the first countries to adopt its legislation, it would send a strong signal to companies around the world that they can and should look to Canada and Canadian companies if they want to develop or acquire security systems. Trustworthy and responsible AIs that respect human rights and protect the well-being of users. »