Cameron F. Kerry “The Framework for Global Law and Accord on AI and Digital”
December 12, 2021
Boston Global Forum
It is a pleasure today to honor Andreas Norlén for his work as Speaker of the Swedish Riksdag in hammering out the formation a coalition government. He has been honored across Europe (DE, Spain, Italy), and it is fitting that he is now being honored in the United States.
Sweden has long been held up as a paragon of social democracy and progressive policies. I remember when it was an opposition research point that Mike Dukakis took reports on Swedish child care policy as beach reading. Sweden has also been a global technology leader – with companies like Ericsson and Spotify, and people like Goran Marby at ICANN.
AN: “Like no other.” The political currents Andreas Norlén faced are evidence that the challenges of fragmentation and polarization, nativism, authoritarianism, and disinformation politics are present even in the most democratic countries on earth.
This makes it all the more important that democratic countries band together. Artificial intelligence, e-commerce, malicious online content, election interference, and cyberthreats are all global issues magnified by network effects that no one country – even the United States – can manage by itself.
The issues are broad and complex, and there is no one forum capable of addressing all the issues. Effective policymaking on challenges enabled by digital technology will many different avenues, a network of networks in parallel to the communications systems that link us.
The Boston Global Forum is one node of the wide and diverse network needed. This week’s Democracy Summit is another. The week before last, I attended the Future Tech Forum, where the UK government convened G7 members with ten other democratic countries that are global or regional leaders in digital technology to explore issues in technology policy and development emerging in the next 5-10 years and their implications for public policies.
Packets of progress can travel by diverse paths across these many networks and assemble into a recognizable whole.
AI has been a focus of the work the BGF and the Michael Dukakis Institute. AI is a critical path for cooperation for several reasons. First, it is essential to successful development. In the 21st century, most scientific research involves international collaboration. But this is particularly the case for AI research and development because it crosses institutions, disciplines, and borders because of the scale of data, compute power, skills needed, and relies heavily on open source software.
Cooperation also presents a unique opportunity for collaboration because every country is still in early stages of policymaking. That makes it possible to align approaches in ways that can avoid divergence of the kind that is causing barriers to the free flow of data between the EU and the U.S. and present a coherent counterpoint to China’s techno-authoritarianism, prevent a race to the bottom.
Since the Policy Lab in September, along with Brookings Institution colleagues and partners at the Centre for European Policy Studies in Brussels, I released a comprehensive report on international cooperation in artificial intelligence policy and development. The report focuses on 15 concrete steps to put cooperation into practice.
It starts with a simple proposition, that likeminded governments “commit to considering international cooperation in drafting and implementing AI policies.” At the UK’s Future Tech Forum, the UK government gave an example of our recommendation in practice: it asked the participating countries to comment on its proposed Online Safety Act. This is an unusual step—and one that needs to become a norm in digital policymaking.
Much of the report focuses on mechanics of regulation, risk assessment, and standards development. Recall how the internet was developed – researchers like Tim Berners-Lee and organizations like WWW Forum, SDOs, ICANN — stakeholders and technicians – hammered out technical standards and systems. These are an example of the kind of policymaking that needs to happen with regard to AI policy and standards – transnational, multistakeholder, public-private collaboration.
One difference from the technical development of the early internet is that the impact is many times broader, and some multinational corporations have acquired much great power. These companies are essential participants in the policymaking but cannot be the ones to make the rules.
There is one other concrete step for international collaboration on AI I want to mention today. It is that likeminded government commit to collaboration on research and development on projects to harness AI for public goods of global importance, like earth observation for climate change, more powerful computation for privacy-protecting ways of sharing data, or monitoring of public health.
Such projects are ways not only to work together on solving important problems for humanity, but also of working through issues like ethics and data governance in the context of specific uses cases where there is strong motivation to get the job done. It is a challenge but, as President Kennedy said about going to the moon, we should do this “not because [it is] easy, but because {it is] hard… that goal will serve to organize the best measure of our energies and skills.”
* * *
AI is just one domain of digital policymaking where the U.S., its allies, and likeminded countries need to work in concert and align their policies. The price of not doing so is divergence and fragmentation. We already in effect have bifurcated internets on this earth, China’s and a shared internet. We don’t need further forking to diminish global network effects that benefit our planet.
I am grateful to BGF for its work to bring leaders around the world together to build cooperation and trust, and to Andreas Norlén for his example of the painstaking and patient building it takes to forge consensus out of disparate positions. These are the tasks ahead.
Download the speech is here