Question Period Note: RESPONSIBLE ARTIFICIAL INTELLIGENCE IN THE GOVERNMENT OF CANADA

The Government of Canada is exploring opportunities to responsibly use artificial intelligence (AI) to improve services for Canadians.

• Canada is the first national government to take concrete steps in defining how AI should be responsibly governed in public service delivery.
• While AI offers promise, we must ensure it is implemented responsibly, governed with clear values, ethics and laws and in accordance with human rights obligations.
• The Treasury Board of Canada Secretariat published a Directive on Automated Decision-Making and an Algorithmic Impact Assessment tool to ensure that AI is used in a responsible manner within the Government of Canada.
• These efforts focus on accountability and transparency, ensuring that all Canadians understand how we are using AI in our programs and services

Artificial Intelligence (AI) can be described as information technology that performs tasks that would ordinarily require biological brainpower to accomplish, such as making sense of spoken language, learning behaviours, or solving problems.

Canada, recognized for its strong talent base in AI development, is among the top five countries in the world in terms of AI experts with over 1,487 PhD-educated researchers.

The Government of Canada has demonstrated its preoccupation with the responsible use of AI through a number of international and domestic AI initiatives.

Domestic AI Initiatives

Over the last three years, the Government of Canada has invested strategically in the Canadian AI ecosystem. Successive budgets funded $125 million toward the Pan-Canadian AI Strategy and $950 million for the Superclusters initiative which are generating cutting-edge research, investment, and talent in Canada. These initiatives are managed by Innovation, Science and Economic Development Canada (ISED).

Announced on May 14, 2019, the Advisory Council on Artificial Intelligence advises the Government of Canada on building Canada's strengths and global leadership in AI, identifying opportunities to create economic growth that benefits all Canadians, and ensuring that AI advancements reflect Canadian values. The Advisory Council is a central reference point to draw on leading AI experts from Canadian industry, civil society, academia, and government.

Government of Canada Directive on Automated Decision-Making

The Government of Canada has also been a leader in the development of guidance for AI use by the public sector:

• Canada’s Directive on Automated Decision-Making, launched in March 2019, seeks to ensure that Automated Decision Systems are deployed in a responsible manner that reduces risks to Canadians and federal institutions, and leads to more efficient, accurate, consistent, and interpretable decisions made pursuant to Canadian law.

• The Directive includes a set of guiding principles which will help ensure departments use AI effectively and ethically:
o Understand and measure the impact of using AI by conducting an Algorithmic Impact Assessment.
o Be transparent about how and when we are using AI.
o Provide meaningful explanations about AI decision-making, while also offering opportunities to review results and challenge these decisions.
o Be as open as we can by sharing source code, training data, and other relevant information, all while protecting personal information, system integrity and Canada’s national security.
o Provide sufficient training to ensure that the employees developing and using AI solutions have the skills needed to make government services better.

• The creation of this Directive was done in collaboration with academics, industry leaders, civil society and other governments to identify and mitigate various risks associated with using AI systems.

• The Algorithmic Impact Assessment (AIA) tool, launched in May 2019, is a questionnaire designed to help government departments assess and mitigate the risks associated with deploying automated decision systems. The AIA also helps identify the impact level of automated decision systems under the Directive on Automated Decision-Making to ensure that the mitigation measures are proportional to the risks. The AIA was also developed in the open and is available to the outside world for sharing and re-use under an open license.

International AI Initiatives

In June 2018, Canada and France announced their intention to create an international initiative on AI. This initiative, now named the Global Partnership on Artificial Intelligence (GPAI), will bring together many of the greatest AI experts globally to support and guide the responsible adoption of AI that is human-centric and grounded in human rights, inclusion, diversity, innovation and economic growth. Canada and France received expressions of interest on GPAI from Germany, Italy, India, Japan, the United Kingdom, New Zealand, and the European Union.

In 2019, Canada participated in the development and adoption of the OECD Recommendation on AI, including the OECD Principles for Trustworthy AI, which serve as the basis for the G20 AI Principles introduced during Japan’s 2019 Presidency and endorsed by Leaders.
TBS is actively engaged with an international group of world-leading digital nations, the “Digital Nations”, in a thematic group on AI to guide governments in their implementation of digital services based on these shared values. Canada is joined in the Digital Nations by Denmark, Estonia, Israel, Mexico, New Zealand, Portugal, Republic of Korea, United Kingdom and Uruguay.

KEY FACTS • The Government of Canada’s Directive on Automated Decision-Making was introduced in March 2019 and comes into force on April 1, 2020 and encourages responsible use of automation in delivering services to Canadians.
• This Directive is supported by the Algorithmic Impact Assessment (AIA) tool to help officials assess and mitigate the risks associated with the use AI within government.