Welcome to AI Policy 101: a new series from Politics + AI that will teach you the fundamentals of artificial intelligence (AI) policy.
This introductory article provides an overview of the field, an explanation for the sudden flurry of national AI strategies, and a breakdown of what AI policy entails. It concludes with a set of key takeaways and a list of further readings.
In the coming weeks, I will share five additional articles that provide a deep dive into key aspects of AI policy. They will cover: (1) basic and applied scientific research; (2) talent attraction, development, and retainment; (3) industrialization and private sector uptake; (4) ethics; and (5) data and digital infrastructure.
Without further ado, let’s begin!
What in the world is AI policy?
First, a definition:
AI policy is defined as public policies that maximize the benefits of AI, while minimizing its potential costs and risks.
From this perspective, the purpose of AI policy is two-fold. On the one hand, governments should invest in the development and adoption of AI to secure its many benefits for the economy and society. Governments can do this by investing in fundamental and applied research, the development of specialized AI and “AI + X” talent, digital infrastructure and related technologies, and programs to help the private and public sectors adopt and apply new AI technologies. On the other hand, governments need to also respond to the economic and societal challenges brought on by advances in AI. Automation, algorithmic bias, data exploitation, and income inequality are just a few of the many challenges that governments around the world need to develop policy solutions for. These policies include investments into skills development, the creation of new regulations and standards, and targeted efforts to remove bias from AI algorithms and data sets.
It is important to note that AI policy is not just the use of AI to improve the effectiveness of government policy or reduce costs. As we will soon see, this is just one of many areas of AI policy.
Why the sudden interest?
Since the beginning of this year, Denmark, France, the UK, the EU, South Korea, and India have all released national strategies to promote the use and development of AI. They join Canada, Japan, Singapore, China, the UAE, and Finland, who all released similar strategies in 2017.
What can explain this sudden interest? Part of the story is a simple proof of concept for AI. In the past six years, computers, powered by AI technologies, have learned how to speak and translate the world’s languages, recognize faces and objects, and even play complex video games. Our favourite services, such as Netflix and Google Search, are now dependent on AI algorithms, while sectors as diverse as transportation and healthcare are set to be fundamentally transformed in the coming years. Simply put, governments now recognize the disruptive impact of AI and want to get ahead of it.
But beyond this technological development is a story of competition. It has become clear that the demand for AI talent far outweighs the available supply. According to a study by Element AI, there are only 22,000 PhD-educated AI researchers in the world — 40% of whom are concentrated in the US. As a result, to train domestic talent and attract international talent, countries are rushing to develop AI Master and PhD programs, short-term training initiatives, massive open online courses, and scholarships and fellowships. Almost every recent national strategy includes some combination of these initiatives to attract, retain, and develop AI talent.
Likewise, governments are also trying to win the global race for AI investment. The UK’s AI Sector Deal is a perfect example. In April, the British government announced a number of new initiatives to establish the UK as a leader in the AI revolution, including a new R&D tax credit, a national retraining scheme, additional funding for STEM education, a national centre for data ethics, and improvements to public digital infrastructure. In return, over 50 companies announced £300 million in private sector investment. The UK is not alone in this effort: France’s strategy included a multi-million dollar commitment to AI startups and industrial projects, while China recently announced a $2 billion AI research park to house up to 400 companies.
Finally, governments are also trying to get ahead of the new challenges brought on by AI. The most widely debated challenge is the future of work and whether robots will automate 15 or 50 percent of jobs. However, recent stories such as the Cambridge Analytica data scandal, Google’s eerily accurate voice assistant, and Amazon’s Rekognition technology have demonstrated to the public the ability of AI to erode democracy, trust, and civil liberties. The more comprehensive national strategies have begun to tackle these issues.
What are the key aspects of AI policy?
AI policy changes from country to country. Depending on a country’s national strengths and weaknesses, a government will choose to focus on different aspects of AI policy. Finland, for instance, wants to lead the world in the application of AI technologies, while Canada wants to be the global leader in AI research and training. The United States has taken a free-market approach to AI policy, while China has implemented a comprehensive, nationwide approach.
Despite these differences, AI policy can essentially be broken down into the following 10 categories:
1. Basic and Applied Research: To achieve new breakthroughs in AI theories, technologies, and applications, governments need to provide funding for basic and applied research. This includes both research grants and the creation of new research institutions. Example: the UK’s Alan Turing Institute.
2. Talent Attraction, Development, and Retainment: To conduct R&D in AI and deploy AI solutions in the public and private sectors, countries need a supply of skilled AI talent. Example: Canada’s CIFAR Chairs in AI Program.
3. Future of Work and Skills: Advances in AI will both create and destroy jobs. To ensure that workers have the skills to compete in the digital economy, governments need to invest in STEM education, national retraining programs, and lifelong learning. Example: Denmark’s Technology Pact.
4. Industrialization of AI Technologies: AI has the potential to fundamentally transform multiple sectors and drive growth for decades to come. To encourage private sector uptake, governments are investing in strategic sectors and developing AI ecosystems and clusters. Example: Japan’s Industrialization Roadmap.
5. AI in the Government: Likewise, governments are experimenting with ways to encourage the uptake of AI in the government. With the help of AI, it is possible to reform the public administration and make policy more effective. Example: UAE’s Ministry of Artificial Intelligence.
6. Data and Digital Infrastructure: Data is central to the ability of AI to work. As a result, governments are opening their datasets and developing platforms to encourage the secure exchange of private data. Example: France’s Health Data Hub.
7. Ethics: Concerns over algorithmic bias, privacy, and security have raised a number of ethical debates. To mitigate harm, governments are looking to develop ethical codes and standards for the use and development of AI. Example: The EU’s Draft AI Ethics Guidelines.
8. Regulations: Every country is grappling with the question of whether (and how) to regulate AI. Currently, governments are focused on regulations for autonomous cars and autonomous weapons. Example: Germany’s Ethics Commission on Automated and Connected Driving.
9. Inclusion: AI can both improve and worsen inclusion. Used properly, AI can bolster inclusion and help address complex societal problems such as poverty and hunger. Used improperly, AI can reinforce discrimination and disproportionately harm women and minorities. Example: India’s #AIforAll Strategy.
10. Foreign Policy: Geopolitics, development, and trade will all be affected by advances in AI technologies. To address ethical concerns and develop global standards, countries are beginning to consider mechanisms for the global governance of AI. Example: China’s Global Governance of AI Plan.
- AI policy is about maximizing AI’s many benefits for our economy and societies, while minimizing its risks and harms.
- Technological advancement in AI can only partially explain the sudden interest in AI policy. Governments are also keenly aware of the limited supply of AI talent and investment and are trying to get ahead of the new challenges caused by AI.
- Governments in all regions of the world are experimenting with AI policy. Currently, there is no best practice since the field is so new. However, AI policy can be broken down into 10 categories: basic and applied research; talent attraction, development, and retainment; future of work and skills; industrialization of AI technologies; AI in the government; data and digital infrastructure; ethics; regulations; inclusion; and foreign policy.
- Malli, Nisa, Melinda Jacobs, and Sarah Villeneuve (2018). “Intro to AI for Policymakers: Understanding the shift.” Brookfield Institute for Innovation + Entrepreneurship.
- Furman, Jason (2016). “Is This Time Different? The Opportunities and Challenges of Artificial Intelligence.” Speech at the 2016 AI Now Conference.
- Dutton, Tim (2018). “An Overview of National AI Strategies.” Politics + AI, Medium.
— — — — — — — — — — — — — — —
Tim Dutton is an AI policy researcher based in Canada. He is the founder and editor-in-chief of Politics + AI. He writes and edits articles for Politics + AI’s Medium page and provides contract work to governments and companies looking to learn about the emerging political risks and opportunities of AI. You can follow him on Twitter and connect with him on LinkedIn.