Policy Briefs

Overview of B20 AI Platform

Artificial Intelligence (AI) is a revolutionary technology that has the capacity and transformative power to dramatically impact the functioning and efficiency of global governance, commerce, and culture. As defined by Google, AI refers to technologies that enable computers to perform advanced functions including the ability to comprehend language and text, analyze and visualize data, and synthesize and generate insights and recommendations.

The potential of AI has dominated policy discussions and investment decisions over the past two years. Global AI deals have grown by 24% year-on-year based on sustained investor resilience and confidence in the sector¹ with the average AI deal size increasing by 28% to USD 23.5 million. Moreover, AI investment has not been concentrated in few markets, but has been widespread with Q3 2024 attracting 566 deals amounting to USD 11.4 billion for US based AI companies, 279 deals worth USD 2.8 billion in Europe, and USD 2.1 billion for Asia based AI companies².

Governments have moved to regulate and develop frameworks for the development of Safe AI to prevent the technology from negatively impacting society. The European Union passed the AI Act in June 2023, which establishes a technology-neutral, uniform definition for AI and ensures that AI systems in the European Union are safe, transparent, traceable, non-discriminatory, and environmentally friendly. Furthermore, the Act imposes different regulations on AI models dependent on their perceived risk levels.

Brazil, which recently hosted the B20 Summit in 2024, is currently debating a proposed law titled Bill N. 2338/2023 that will regulate AI and follows the European Union’s risk- based approach with bans on AI models that carry excessive risk and increased regulation on AI models that are identified as being high-risk systems. Brazil’s proposed AI law expands on the EU’s regulations by imposing specific obligations on the proprietors of general-purpose AI systems.

China too introduced regulations on AI in 2021 that have since been expanded to include regulations on algorithm recommendations, deep synthesis, and generative AI.

B20 AI Platform

Recognising the transformative power of AI, and the role that the technology would play in shaping global economic and governance trends in the coming years, B20 Brasil developed the B20 AI Platform. As AI was a central theme in the recommendations of several B20 Brasil Task Force papers as well as the conclave itself, the AI platform was developed as an aggregated repository of recommendations on the field of AI³ .

The first category under the B20 AI Platform includes foundational actions that G20 countries could take to promote the development of safe, advanced AI. First, B20 Brasil recommends that Digital Inclusion is prioritised to ensure and achieve connectivity between individuals and businesses. Specifically, B20 Brasil suggests that modern regulations as well as public-private partnerships could facilitate greater accessibility and resilience in ICT (Information and Communication Technologies) across regions. Second, the B20 AI Platform notes that G20 countries could harmonize cybersecurity standards, coordinate cybersecurity action, and support Free Flow of Data with Trust to protect individuals and organisations and promote digital trust and innovation.

Next, B20 Brasil Task Forces recommend several potential actions that G20 countries could take to harness the transformative power of AI:

• • First, B20 Brasil recommends that countries work together to strengthen international collaboration and develop pro-innovation, risk-based frameworks for the development and responsible governance of AI.
• • Second, the importance of implementing bias-free AI that considers diversity and inclusion (D&I) and underrepresented groups was emphasised.
• • Third, the report encourages the development of transparent and clearly communicated AI models as well as codes of conduct on AI.
• • Fourth, B20 Brasil calls for the development of AI clusters and ecoystems that support collaboration between research centres, government agencies, and startups to advance AI capabilities.
• • Fifth, B20 Brasil advocates for reducing digital divide through upskilling and reskilling programs. Specifically, it advocates for the implementation of vocational training programs as well as improved access to basic and higher education in digital skills through, for example, public-private partnerships, technological residencies, and micro-credentials among other initiatives.
• • Sixth, B20 Brasil recommends that G20 nations promote labor policies that improve productivity as well as improve the resilience and innovative capacity of the workforce in advanced technological fields such as AI.
• •Finally, seventh, B20 Brasil recommends that G20 nations promote digitalization in international trade. It advocates for simplification and harmonization of customs processes, regulatory cooperation, and digitalization of trade processes through AI particularly in the risk profile assessment in the financing application process.

Conclusion

Based upon these policy roadmaps, G20 countries can work together or independently to implement initiatives within their respective countries. By harnessing the transformative power of artificial intelligence, G20 countries can improve their productivity and streamline agricultural, digital, and industrial processes.

Moreover, G20 nations must act rapidly to safeguard national and global security from high-risk AI models that could spread misinformation or be potentially leveraged to propagate and influence society in a negative manner. Therefore, both industry and governments must work in tandem to advance artificial intelligence and the implementation of artificial intelligence in the commercial sector. At the same time, they must work on developing modern regulations that facilitate the advancement of artificial intelligence while also preventing dangerous externalities.

Artificial Intelligence will transform the capabilities of knowledge workers and dramatically increase the productivity and capacity of these workers to develop content. However, at the same time, these models could produce significant risks of misinformation and bias, which could cause significant societal harm. Therefore, aligning with B20 Brasil Task Force recommendations, governments should significantly invest in promoting digital skills and familiarity with artificial intelligence to make citizens aware of the technology and more capable to leverage it positively.

Note: