Antara Vats and Nikhila Natarajan, G20.AI: National Strategies, Global Ambitions, July 2022, Observer Research Foundation and Observer Research Foundation America.
I. Executive Summary
“If a machine is expected to be infallible, it cannot also be intelligent.” – Alan Turing[1]
Many countries have launched their national strategy documents on Artificial Intelligence (AI) in recent years; in the last five years alone, more than 60 did so, following Canada which was the first to release its document in 2017. In Europe, 20 member states of the European Union (EU), and Norway, had published their national AI strategies by 2021.
This report,G20.AI: National Strategies, Global Ambitions, gives an overview of the AI strategies of G20 countries. It is the second in a series of ORF reports on AI, the first of which was published in 2018 and laid out the national strategies of 12 countries that had released theirs at the time.[2]It is vital to understand G20 approaches to optimising the benefits of this transformational technology, as India prepares to take the G20 presidency in December 2022.
Fifty years after the internet was born, public outcry against surveillance technologies is forcing governments to rewrite the norms of the public square in the digital, networked era. A rethinking of the economics is underway: How much automation is politically and economically sustainable? Meanwhile, algorithms are toying with social faultlines; those with access to cryptocurrency and a 5G-smartphone are challenging central banks and bankers; and angry short sellers banding together on Reddit are threatening the clout of hedge funds. A great transition is afoot, and at the heart of this churning is AI.
As countries and companies conduct AI research and deploy the technology in the public sphere, who is to police it and how? The past year has laid bare the weaknesses of, and the threats to digital democracy. The demand for tech companies to ensure algorithmic fairness, accountability, transparency, and ethics has become louder. Large, private-owned platforms designed and anchored in Anglocentric worldviews push the kind of free speech absolutism that is in conflict with laws in most democracies. Moreover, these same companies have long had a stranglehold on the AI agenda,[3]and they must be scrutinised.
The influence of these companies covers a massive terrain: Facebook alone has 3 billion users; Twitter has 300 million; and many individual “influencers” have millions of followers. As the populations of social media users across the globe expand exponentially, regulators are scrambling to catch up. Indeed, a clash of norms is upon the world. This is very different from how it was in the 1990s, when the internet first evolved from being an obscure research network into something that millions of people used in their everyday lives. At that time, the public was told that more participation in the digital world would strengthen democracy. Thirty years since, certain countries have succeeded in using the internet as a means to interfere with the elections of other sovereign countries; there has been a rise in the incidence of State-sponsored malicious cyber activity; tech companies have begun to stockpile and sell data, provoking privacy concerns; and social media platforms have become potent vectors for the spread of misinformation and disinformation, turning people against one other and dividing societies.
A Deloitte global survey of 500 government leaders in 2021 found that 92 percent of respondents at the federal level, 95 percent at the state level, and 84 percent at the local level are of the view that AI is “mission critical” over the next five years.[4]At least half of those surveyed listed gaps in skills as a crucial reason for the inability of governments to utilise AI applications in the most effective manner.
Indeed, many lower- and middle-income economies are taking ambitious steps in AI innovation. Geographies like India, for example—where the scale of societal problems is massive and the solutions are complex—are witnessing the proverbial snowball effect: it is not any one line item that tips the scales but simultaneous, cutting-edge developments. The results are remarkable: A Nasscom report[5]lists India as a promising nation for innovative technology. India has filed over 6,000 applications for AI patents over the last decade, more than 94 percent of which were in the past five years alone.
To be sure, the politics of this process is unavoidable, as it has been in earlier innovation cycles. The shifts in technological expertise from the fabled locale of Silicon Valley to formerly colonised swathes of the Global South are combining to create an undeniable force. What is similar across nation states trying to come up with strategies for artificial intelligence? For one, the careful dance around definitions of ‘fairness’ and the nebulous construct of AI ethics.
The contours of this story are changing in national and cultural contexts, but the idea that technological systems normally preserve existing hierarchies and power structures continues to hold true in more ways than what are immediately obvious. Where do countries stand on the promise and the peril of AI? How are they articulating it in government agendas? Observer Research Foundation presents Round Two of a global snapshot:G20.AI: National Strategies, Global Ambitions.
II. Rationale And Methodology
The substantive frame of this report is limited to the national AI strategies of G20 countries. (Table 1 lists the current status of national strategies and approach documents as of March 2022.) This report analyses the domestic agenda and capabilities of each nation in the context of its global priorities, using a combination of textual analysis and cross-referencing a common pool of AI strategy reports. Throughout the analysis, US-China relations serve as a backdrop.
Table 1. National AI Strategies/Approach Documents of G20 economies
National AI Strategy/Approach Document
Year of Release
Plan Nacional de IA Gobierno de Argentina – IALatam[6]
Decree of the President of the Russian Federation on the Development of Artificial Intelligence in the Russian Federation – Centre for Security and Emerging Technology[19]
“Coordinated Plan” on Artificial Intelligence 2021 Review | Shaping Europe’s digital future[27]
2021
In some countries where one incumbent government had published a strategy document and the successive government is shaping its own version, this report refers to the entire combination of those documents, presuming them to speak of the nation’s strategic bent. In the United States (US), for example, the Donald Trump administration released the document,Artificial Intelligence for the American People[28]in 2019 and theNational AI Initiative Act of 2020[29]came into force a fortnight before Joe Biden took over the presidency. Within a month, the National Security Commission on Artificial Intelligence issued a 756-page blueprint outlining steps that the US Government should take to “win” in the AI era. In this case, this analysis considers the totality of the recommendations across the different documents.
This study utilises four parameters that thread together all the G20 AI documents and are fundamental to building an AI ecosystem: Research and Development (R&D); Skills; ICT Infrastructure; and Data Ecosystem. The national strategy documents share these common elements, allowing the authors to compare the level of specificity that each country accords to each pillar, their definitional boundaries, and the domestic priorities that may be animating the different approaches.
This analysis also uses the following resources for context and comparison:Building an AI World: Report on National and Regional AI Strategies[30](Canada); Global AI vibrancy tool[31](Stanford Institute for Human-Centred Artificial Intelligence); theGlobal AI Strategy Landscape[32](Holon);The AI Policy Observatory[33](OECD) that tracks 700 AI policy initiatives from 60 countries, territories and the EU; theGlobal Innovation Index[34](World Intellectual Property Organisation); AI Index[35](Stanford Institute for Human-Centred Artificial Intelligence); AI Readiness Index[36](Oxford Insights); Investment Tracker[37](Georgetown Centre for Security and Emerging Technology); Data Protection and Privacy Legislation Worldwide[38](United Nations Conference on Trade and Development); Network Readiness Index[39](Portulans Institute); and Global Talent Competitiveness Index (INSEAD).[40]
The strength of each country’s AI strategy is illustrated on a radar plot and the surface area of the plot indicates each country’s relative strength. The farther out the nodes for each parameter are on the plot, the stronger the strategy, and vice versa.
Like any analysis that involves a close reading of documents issued by governments, this report is limited in qualitative interpretations of technological, geopolitical, and strategic details. Government strategies may not reveal the full gamut of applications that are under consideration or even the shifts in strategy based on new information that is available to the government but not yet in the public domain. Despite these limitations, this report is part of ongoing efforts by the Observer Research Foundation to understand the broad trajectory in which a transformational technology is affecting governmental actions.
Antara Vats is a Junior Fellow at Observer Research Foundation’s Centre for Security, Strategy and Technology.
Nikhila Natarajan is Senior Programme Manager for Media and Digital Content with ORF America.
[1]Lecture to the London Mathematical Society, 20 February 1947. Quoted in B. E. Carpenter and R. W. Doran (eds.), A. M. Turing’s Ace Report of 1946 and Other Papers (1986); Alan Turing has played a pivotal role in laying the foundations of artificial intelligence and theoretical computer science.
[2]Samir Saran, Nikhila Natarajan and Madhulika Srikumar,In Pursuit of Autonomy: AI and National Strategies,India, ORF, 2018,
[34]Soumitra Dutta, Bruno Lanvin,Lorena Rivera León and Sacha Wunsch-Vincent,Global Innovation Index 2021: Tracking Innovation through the COVID-19 Crisis, World Intellectual Property Organisation, 2021,
[35]Daniel Zhang, Saurabh Mishra, Erik Brynjolfsson, John Etchemendy, Deep Ganguli, Barbara Grosz, Terah Lyons, James Manyika, Juan Carlos Niebles, Michael Sellitto, Yoav Shoham, Jack Clark, and Raymond Perrault,The AI Index 2021 Annual Report, AI Index Steering Committee, Human-Centred AI Institute, Stanford University, Stanford, CA, March 2021,https://aiindex.stanford.edu/wp-content/uploads/2021/11/2021-AI-Index-Report_Master.pdf;Daniel Zhang, Nestor Maslej, Erik Brynjolfsson, John Etchemendy, Terah Lyons, James Manyika, Helen Ngo, Juan Carlos Niebles, Michael Sellitto, Ellie Sakhaee, Yoav Shoham, Jack Clark, and Raymond Perrault,The AI Index 2022 Annual Report, AI Index Steering Committee, Stanford Institute for Human-Centred AI, Stanford University, March 2022,https://hai.stanford.edu/research/ai-index-2022.
[37]Zachary Arnold, Ilya Rahkovsky and Tina Huang,Tracking AI Investment: Initial Findings From the Private Markets,Georgetown Centre for Security and Emerging Technology, 2020,
[40] Bruno Lanvin and Felipe Monterio,The Global Talent Competitiveness Index: Global Talent in the Age of Artificial Intelligence,France, INSEAD, 2020,
The views expressed above belong to the author(s). ORF research and analyses now available on Telegram! Click here to access our curated content — blogs, longforms and interviews.
Antara Vats is an IIC-UChicago Fellow currently stationed at the Ministry of Electronics and IT Government of India where she works on projects related to ...
Nikhila Natarajan is Senior Programme Manager for Media and Digital Content with ORF America. Her work focuses on the future of jobs current research in ...