Advocates urge full public consultation on controversial AI legislation

Call comes after Industry Minister publishes 38-page package of proposed amendments midway through Committee hearings

DECEMBER 14 2023 — Today over 25 leading civil society organizations, experts, and academics released an open letter to the House of Commons Industry Committee, urging them to hit the reset button and fully scrutinize the government’s controversial Artificial Intelligence and Data Act (AIDA). They recommend a full public consultation and redrafting of AIDA, starting with splitting AIDA from the other parts of Bill C-27, which deal with unrelated privacy matters, so that it can be subject to the careful democratic scrutiny it requires.

Highlighting what they describe as ISED’s mishandling of “a process biased heavily toward narrow industry interests”, the signatories call on MPs to ensure that any future public consultation is not stewarded exclusively by ISED. The call comes two weeks after Industry, Science, and Economic Development (ISED) Minister François-Philippe Champagne published a beefy, 38-page package of proposed amendments to AIDA, rivaling in size the text of the original bill, and fundamentally altering its shape and implications. 

The signatories express grave concern that many relevant expert witnesses and civil society organizations will be denied the opportunity to provide input and testimony on the Minister’s  fundamentally revised version of AIDA, given that its publication comes half way through the Industry Committee hearings. They highlight the risk of a radically reshaped AIDA passing into law without proper consultation or informed discussion, and argue that this is “inconsistent with the norms one expects from a democracy.”

AIDA has been subject to fierce criticism since it was first introduced in June 2022, without the public consultation process that typically precedes similar legislation. Key concerns include:

• Its failure to recognize fundamental human rights, such as privacy and free expression, that AI must be designed and deployed to respect;

• Its lack of any independent oversight or enforcement, instead placing its AI & Data Commissioner under the authority of the Industry Minister, who sponsors and subsidizes the AI industry;

• Its failure to address the societal level risks of AIDA, including risks to marginalized communities;

• Its light-touch, mark-your-own-homework approach to regulating the AI industry, which is inconsistent with the serious potential dangers of AI;

• Its failure to consult the public, instead heavily prioritizing industry input.

Over the past two years, more than 14,000 signatures and letters have been sent to government officials calling for strong action to address the impacts of AI and facial recognition. More than 30,000 signatures have been collected by OpenMedia petitions calling for new privacy laws in Canada, and more than 17,800 messages have been sent to the government calling for enhanced personal privacy protections. 

Quotes

“AI technology may fundamentally transform our society in the years ahead – and that makes it critical that we regulate it effectively, not hastily. Today we’re urging the government to go back and do their homework, ensuring their legislation is stress tested by a wide public consultation and dedicated parliamentary consideration, not shoe-horning it into a fundamentally distinct privacy bill.”

• Matt Hatfield, Executive Director at OpenMedia

“AIDA is deeply flawed in its substance and the process by which it has been developed. It has been shaped primarily by business interests to the exclusion of those experiencing its harms or potential risks. It needs to be re-drafted properly.” 

• Andrew Clement, Professor emeritus, University of Toronto

“Excluding private sector AI tech developed for government intelligence, defence and national security purposes from any form of regulation means a free pass for some of the most potentially harmful AI tools. If the government is serious about protecting the rights of people in Canada, AIDA isn’t up to task.” 

- Tim McSorley, National Coordinator, International Civil Liberties Monitoring Group

“Data shows that Canada has a public trust problem with AI. This patchwork legislation embedded in a problematic process cannot fix that trust gap. It’s time to reset and consult on a coherent bill to create a solid foundation for the trustworthy AI innovation we deserve.”

• Brenda McPhail, Acting Executive Director, Master of Public Policy Program in Digital Society, McMaster University