Page 29 - AI for Good - Impact Report

AI for Good

is anticipated that this could lead to another "Brussels effect," similar to the impact seen with
the GDPR. Therefore, while the EU AI Act will have a global impact on businesses, it is also likely
that some governments will incorporate elements of the EU's law into their legislation. Key
aspects of the EU AI Act that may be adopted by other countries or regions include the risk-
based approach to AI applications, the protection of citizens, obligations for general-purpose
AI, as well as transparency and cybersecurity measures. Ultimately, in the context of the Act’s
extended timeline for full applicability, there is potential for regulatory "cherry-picking" by other
governments. As the EU demonstrates its capacity to implement successful AI governance, other
nations might selectively adopt certain provisions that align with their domestic priorities while
bypassing others. This selective adoption could be influenced by the demonstrated effectiveness
of the EU’s implementation. As a result, while the EU AI Act sets a comprehensive framework, its
influence may lead to a varied global regulatory landscape where different governments tailor
their AI governance frameworks by integrating specific elements of the EU's model.

National level

Additionally, many governments have started to develop ethical frameworks to address these
matters. These approaches range from regulations, to codes of conduct or AI strategies. Efforts
are diverse yet share common goals of ensuring ethical use and mitigating risks. Across these
diverse approaches, there is a clear consensus on the need for frameworks that ensure AI systems
operate transparently, fairly, and responsibly. These regulations and strategies collectively
reflect a global commitment to addressing the ethical and societal impacts of AI, aiming to
foster technology that aligns with core values of human rights and public trust.

However, the implementation of these frameworks is often inconsistent and varies greatly
between countries. For instance, while some countries like Singapore and the United
102
Kingdom have developed comprehensive ethical frameworks for AI, others are still in the
103
early stages of this process. Even when ethical frameworks exist, there can be challenges in
enforcing them and ensuring compliance.

China's regulatory landscape reflects a focus on comprehensive oversight and national security,
particularly in ensuring the safe use of AI. This focus is underscored by China’s recent framework
addressing the security governance of AI. The framework is designed to ensure that AI systems
104
are adaptable and flexible, capable of effectively responding to evolving environments while
maintaining stringent safety standards. It also advocates for a proactive approach to secure and
responsible AI development, prioritizing the identification and management of AI-related risks
through robust technical measures. Further key regulations, such as the Personal Information
Protection Law (2021) and the Data Security Law (2021), emphasize informed consent, data
106
105
protection, and algorithmic fairness. Furthermore, recent regulations like the Gen AI Regulation
(2023) specifically address the challenges posed by GenAI technologies.
107
In Canada, the emphasis is on transparency and accountability in AI systems. The Directive on
Automated Decision-Making (2019) and the draft Artificial Intelligence and Data Act (2022)
108
109
are designed to ensure that AI technologies are developed and deployed responsibly, with
a focus on managing risks and maintaining fairness and include provisions for transparency,
accountability, and the protection of privacy.

Around the world, numerous countries are actively developing and implementing AI strategies to
address the ethical, societal, and regulatory challenges posed by these technologies. According
to the OECD, by 2021, 70 states, including many countries of the Global South, had already
110

19