Digital Policy Hub

Each new release of large language models (LLMs) often comes with claims of both improved performance and enhanced safety. However, there is a lack of standardized safety assessments and a gap in studying these metrics over time. This working paper aims to address this gap by analyzing performance on various standardized safety benchmarks across various LLMs released in the last three years to gauge if they are becoming safer.

The United States and China are innovating stronger military technologies through civil-military fusion. Both states are leveraging converging technologies such as semiconductors and artificial intelligence to strengthen their weapons.

Growing interest in Africa’s digital transformation is being driven by the continent’s youth bulge, its potential as a partner trading on an equal footing and its role in tech geopolitical rivalries. Regional approaches to data governance help enhance interoperability and harmonization of national data regulations, providing African states with more leverage and market share in the digital economy.

There is a pressing need for further research on how cyberwarfare will shape states’ incentives to buy in to international AI coordination regimes. Research in this vein should seek to identify and leverage cybersecurity vulnerabilities that could be used to increase the effectiveness of international coordination efforts.

The rise of digital platforms has transformed communication but also enabled the spread of hate speech and divisive content. Algorithms optimized for engagement have created feedback loops that reinforce polarizing behaviours and narratives, highlighting the tension between user engagement and harm reduction.

Regulating digital technologies is a delicate balancing act of protecting personal data and privacy rights without stifling innovation. This working paper looks at how Sweden and Korea are managing these digital governance challenges and what Canada can learn from these innovation leaders.

Organizations and governments are increasingly using data and artificial intelligence in homelessness management, housing allocation and real estate markets to improve resource matching, predict trends and optimize housing support. These technologies pose serious concerns around privacy, discrimination and bias.

Public health plays an important role in the broader health system by promoting and protecting citizens’ health, often under challenging circumstances, including declining trust, limited resources and complex information ecosystems. Generative artificial intelligence offers potential opportunities for public health to tailor health information to diverse populations at scale, a current challenge of public health programs and service delivery.

This working paper utilizes a feminist policy analysis framework centred around five thematic areas — intersectionality, context, neutrality, control and power — to analyze global initiatives for AI safety governance.

This working paper explores how US-based big tech corporations use their corporate venture capitalist arms to heavily influence the development of emerging technologies by shaping which innovations receive funding.

Temporary foreign workers (TFWs) have been coming to Canada for nearly six decades to do work foundational to the country’s economy. Today, as political sentiments shift to the disapproval of TFWs, it is imperative that the language used to describe them and their work remains respectful.

In Canada, the potential of financial technology to enhance accessibility, diversify offerings and bring new competition to its oligopolistic banking sector is still in its early stages. Consumer adoption of financial technology (fintech) products and services remains relatively low compared with other advanced economies, despite a generally positive perception of fintech services among consumers. New participants in the banking sector offer innovative, competitively priced services, but they face significant barriers.

Outer space resources are available to all, but their supply is limited. As lunar activity and competition for resources increase, so does the need for lunar resource governance. Canada’s extensive experience in the mining sector can help guide a better approach to governing space resources.

Since its formation in 2003, the Canada Border Services Agency (CBSA) has been the only major federal law enforcement organization in the country without an arms-length watchdog agency. The recent passage of the Public Complaints and Review Commission Act addresses this gap in accountability by extending the oversight model of the Royal Canadian Mounted Police to the CBSA.

Countries pursuing liberal internationalist norms can engage with international institutions to promote their interests through norm entrepreneurship. With its established research community in the domain of quantum science and technology (S&T) and a seat at the table in many international institutions, Canada has an opportunity to promote norms of responsible use and international equality in the domains of quantum S&T.

China has been exporting digital cultural products that function as non-state actors, including video games such as Black Myth: Wukong, which incorporates Chinese localized narratives, mythologies and traditional cultures. Through an examination of China’s digital soft power and the underlying tensions between grassroots counter-narratives and state-driven propaganda, this paper attempts to address gaps in Canada’s digital governance related to the Indo-Pacific Strategy.

It is crucial to have an agile model of AI governance anchored in risk-based, rights-based and rules-based principles. Proposed solutions include a universal AI convention enforced by a High Commission for AI and Human Rights, an Intergovernmental Panel on Climate Change-like panel for rigorous research and policy guidance and a global research consortium inspired by the European Organization for Nuclear Research to ensure inclusive, transparent AI development and equitable benefit sharing.

The Canadian Armed Forces is not able to accurately quantify the rate of sexual misconduct in the organization, nor the trends associated with it, despite several cycles of institutional reform in recent decades. Data fragmentation, poor data management, data not fit for purpose and a lack of accountability have obscured organizational knowledge on sexual misconduct, contributing to the “policy graveyard” phenomenon seen in recent attempts for military reform.

Civil society organizations should play a larger role in the artificial intelligence (AI) policy-making process because they offer valuable insights into the societal and ethical implications of AI. These organizations can be empowered by enacting various institutional changes, specifically by connecting them with AI governance bodies such as the Advisory Council on AI, the Office of the AI and Data Commissioner and the AI Safety Institute.

The 2024 US elections underscored Russia’s continued role as an aggressive foreign actor interfering in US democratic processes. Russia employed sophisticated disinformation strategies to influence public discourse and election outcomes. Machine learning analysis of Russian propaganda revealed several key narratives present in the election discourse.

Digital technologies play a crucial role in modern urban life, as digitization levels continue to increase globally. Although quantum-enhanced intelligent transportation systems have the potential to accelerate the efficiency of smart cities, they also come with risks and governance challenges.

In the intensifying global race for advantage in artificial intelligence (AI), Canada, despite its long and illustrious history this area, is beginning to lose its competitive edge. AI policy priorities need to be tightened and focused, if Canada is to have more enduring comparative advantage.

North American nations are expanding the digitalization and incorporation of artificial intelligence at and around their borders amid record numbers of displaced people worldwide. A delicate balance is essential for policy makers, as they conduct cost-benefit analyses of employing sophisticated border technologies, to ensure that internationally protected human rights are upheld, while prioritizing state security and public safety.

Data access and exchange offers benefits but also brings to the fore concerns about privacy rights, ethical implications and the effectiveness of existing regulatory frameworks.

The Fourth Industrial Revolution, or Industry 4.0, is introducing rapid technological advancements that will fundamentally alter wealth distribution, labour automation and the creation or reinforcement of monopolies. While it is not as internationally prominent as those of other Industry 4.0 leaders, Canada is making significant advancements to improve and develop its advanced manufacturing industry, which has the potential to actively participate in developing critical Industry 4.0 technologies and processes.

Information sovereignty exists when people are part of a healthy and culturally appropriate information ecosystem, in which quality and diverse information is available, accessible and stable and it is collected, stored, managed and disseminated with ethics, meeting peoples’ information needs and preferences for an open, inclusive and plural public sphere.

The Artificial Intelligence and Data Act (AIDA), one part of the Digital Charter Implementation Act or Bill C-27, aims to regulate the development and use of artificial intelligence systems by companies operating in Canada. Another part of the bill, the Consumer Privacy Protection Act (CPPA), aims to regulate the collection and use of Canadians’ personal information by these companies. Both AIDA and the CPPA can be modified to better reflect the interests, and satisfy the demands, of non-governmental actors.

Australia, Canada, Ireland, the Netherlands and the United Kingdom have all devised digital regulators forums. These forums increase regulators’ capacity for evolving digital regulations by bringing competition, privacy and communications regulators together to share insights, data and policy approaches toward the regulation of online platforms. The emergence of digital regulators forums can allow for comprehensive digital policy; however, with these forums in early stages, there are questions as to how they will achieve this comprehension.

Artificial intelligence (AI) risk in the research sector will require international governance to ensure the preservation of an independent and reliable global research sector. While many types of governance and safety tools will be necessary, the use of international standardization would be a useful initial phase that could allow for rapid and inclusive governance on which other frameworks could build. International standardization for AI implementation in research institutions could help to build global consensus on preserving and protecting research institutions and their information from AI risk and could help to support other efforts at global AI governance outside of the sector.

Canada has recognized the importance of the next generation of quantum technologies for a variety of applications; however, Canada’s level of public investment remains somewhat modest compared to other leading nations. Increased public investment in quantum science and technology can help secure Canada’s position within its alliance politics, not only because general investment will alleviate allies’ concerns about budgetary allocations, but also because specific quantum technologies are poised to address the unique challenges facing Canada’s key alliances.

Generative AI can be applied in various agricultural extension and advisory services, including farm mechanization, food processing, water management, crop monitoring and livestock management. Most training data sets contain copyrighted works, raising legal questions about their use, especially in agricultural extension services, where training data includes copyrighted images of soil, climate and plant conditions.

Countering climate disinformation requires a multi-faceted approach that includes global information sharing and collaboration, capacity building and technical assistance, regulatory measures and legal frameworks, engagement with technology platforms, and support for independent journalism and public awareness campaigns.

This working paper explores the intricate factors affecting the adoption and diffusion of satellite internet technologies in Sub-Saharan Africa, specifically in Kenya. With a digital divide focus, this paper seeks to understand why rural Kenyans, who constitute 75 per cent of the population, prefer older, less cost-effective broadband technologies over newer, more affordable satellite internet options.

This working paper demonstrates how the Gig Economies Living Lab (GELL) model can be applied as a framework to conduct policy research through a critical feminist approach that centres the micro-level everyday experiences of workers in the gig economy. The author examines the intersection of technology with individuals’ daily work and life practices in relation to their bodies and emotions.

When it comes to central bank digital currencies (CBDCs), this working paper offers a novel perspective by shifting focus from central banks to the crucial role played by governments. It explores the democratic governance required to make an informed decision on whether to issue a CBDC and, if one is to be issued, how to develop and implement it responsibly.

This working paper is the second in a series that looks at how a Gig Economies Living Lab could serve as a collaborative research hub to study the emerging digital economy based on the experiences of gig workers.

This working paper explores what the digital Canadian dollar project can learn from the digital euro and digital pound to cultivate a more user-focused approach, while remaining neutral on the question of its necessity.

Korea and Sweden have shown significant progress in adopting digital technologies as well as in reckoning with the challenges that digitalization brings. In the case of Canada, while progress has been made in some domains, governance of digitalization has been characterized by more jagged priority setting. Overall, there is a distinct lack of policy durability and targeted focus on digital priorities. As Canada confronts digital transformation, there are policy lessons it can learn from the Korean and Swedish experiences.

Expanding climate-smart agriculture relies on deploying artificial intelligence- (AI-) driven digital technologies, such as big data, machine learning and deep learning. Farmers are concerned about sharing their data because they fear it may be mishandled by untrustworthy parties and used against them, particularly as they typically encounter two primary legal challenges in AI-enabled climate-smart agriculture: data ownership and privacy.

Existing governance structures for data and information in Canada-based research institutions are varied and often overlapping. The sector requires binding standards for machine learning/artificial intelligence (ML/AI) deployment, alongside broad strategic planning, the promotion of safe experimentation with ML/AI tools and the development of frameworks for institutions to mobilize and exchange their data.

The current recognized international standards on the right to information must be expanded to include other crucial dimensions beyond accessibility, such as availability, quality, stability, ethics, cultural appropriateness, agency and usability.

Current quantum strategy in Canada does not consider gender-related issues in a robust manner, often promoting inclusive workspaces as a means to fill high-demand roles.

Artificial intelligence (AI) is quickly becoming the weapon of choice in military conflicts. Despite the advantages of AI-based military applications, such as enhanced data analysis and decision making, the risks are substantial, compounded by the lack of laws governing their use.

Climate change is a particularly acute issue for Africa, as the continent’s already fragile environment has been severely impacted. Climate disinformation is an existential risk for many on the continent: misinformation can decrease support for effective mitigation policies and also significantly limit effective adaptation measures.

Decentralized autonomous organizations provide a modern and promising substitute for the conventional Village Savings and Loan Association model by integrating blockchain and smart contract innovation, thereby allowing community-based finance and governance in developing economies in the Global South to be scaled up.

The fast-paced integration of large language models into societal facets such as finance and security introduces a new set of complex ethical, legal and rights challenges, emphasizing the need for a unified international governance framework for artificial intelligence (AI). Such a framework underscores the call for a universal convention protecting humanity’s most fundamental rights and values against unethical and unsafe AI systems.

The dominance of foreign entities in African digital infrastructure and data control threatens national sovereignty and perpetuates inequalities, with African data often exploited by foreign corporations, echoing colonial resource exploitation patterns.

The world of higher education is being transformed by generative artificial intelligence (AI), as instructors, researchers and students begin using this technology for various purposes. If used ethically and responsibly, this technology can enhance teaching, research and learning practices. If used recklessly and unscrupulously, however, it can undermine academic integrity and produce negative educational outcomes.

This working paper delves into strategies to mitigate the negative impacts of deepfakes while simultaneously highlighting the measures that various actors, including states, social media platforms and tech companies, are taking to tackle this issue.

This working paper uses open-source reporting from organizations that experienced cyberattacks, government agency reporting and quantitative analysis using eSentire’s internal data set on cyberattacks detected and responded to in customer environments in North America. This data set demonstrates the value of managed detection and response firms and how they have helped to reduce the cost of cyberattacks to North American organizations by preventing attacks before they have serious financial consequences.

The nascent digital economy in Africa presents both opportunities for indigenous innovation and development as well as challenges that have the potential to worsen expected gains. This working paper gives an overview of the critical challenges and opportunities for developing sustainable data governance frameworks in the region.

Against the backdrop of a more regulated internet and enhanced digitalization, the United Nations is proposing both a Global Digital Compact and a Digital Cooperation Forum ahead of the Summit of the Future. These have the potential to upend much multi-stakeholder internet governance work, shifting how the internet has been governed since its public release in the 1990s.

Big tech corporations play a significant role in developing and maintaining emerging technologies, such as Web3 and metaversal tech, by providing cloud infrastructure services that enable developers to create and deploy these technologies. Effective regulation requires analyzing the behaviour of entities controlling emerging tech, particularly their political andeconomic motives.

With increasing investments in Canada’s artificial intelligence (AI) computing capacity, little attention has been paid to the growing environmental harms that emerge from scaled-up AI developments. To address those harms, this working paper proposes including environmental impact assessments and corresponding limitations on environmentally degrading developments.

Foreign interference is a growing threat, and elections in democratic states are particularly at risk. China has been implicated in various covert political operations targeting the Chinese diaspora in Canada, using platforms such as WeChat and TikTok to influence federal elections.

Kenya’s digital divide is driven by infrastructure issues, high costs, digital illiteracy and cultural norms. Progressive policies should encourage equitable access by lowering taxes and tariffs.

The goal of this paper is to present an intuitive summary of computational learning theory, and its application for analyzing the most popular learning algorithms in machine learning, neural networks.

Despite the adoption of various patent-based mechanisms, the inadequacy of patent systems in incentivizing green technology is well documented and persists. A potential solution to capture more positive externalities associated with green technologies involves the implementation of a patent prize system, also known as a “patent reward.”

While significant research has focused on the economic and technological aspects of central bank digital currency, the governance perspective has not received enough consideration.

Technology-facilitated gender-based violence (TFGBV) is a pervasive aspect of digital societies and requires urgent and innovative solutions, which include the adoption of smart regulations. The EU Digital Services Act is a recent example; however, challenges that may jeopardize its potential for success in addressing TFGBV must be addressed.

Climate change is the quintessential global challenge, while also perhaps the issue that has experienced the most polarization in recent years. New policies and approaches for policy development and implementation will be required to match the alacrity of the proliferating online flows of misinformation and disinformation.

An institutional framework for global artificial intelligence (AI) governance should incorporate lessons from international organizations to guide AI’s ethical development and deployment within and beyond national borders.

This paper presents the proposition for a gig economies living lab model that potentially serves as a hub and space for multi -stakeholder research, data gathering, cooperation, engagement and innovation.

With the surge in generative artificial intelligence (AI) models, voice cloning bots are rapidly surfacing, allowing users to replicate the voice of any artist or celebrity of their choice. This working paper deciphers the use of AI in creative fields and provides some recommendations on policy developments.

Research institutions, for example, research libraries, archives and universities, are an essential part of society that are at high risk for artificial intelligence (AI) adoption within basic systems such as search and cataloguing functions. This working paper surveys existing policies for AI in universities, ethical frameworks for library and archive associations, and existing international and national policies that are relevant to helping to construct a policy framework based on existing norms, practices and values.

Large language models through generative artificial intelligence (AI) offer a unique opportunity to bridge linguistic gaps for international students, assisting non-native English speakers while minimizing universities’ costs. This paper also explores the ethical considerations of their usage and suggests a university policy framework for responsible generative AI use in academic settings.