Connect with us

Science & Technology

Artificial Intelligence and Its Partners

Published

on

Authors: Oleg Shakirov and Evgeniya Drozhashchikh*

The creation of the Global Partnership on Artificial Intelligence (GPAI) reflects the growing interest of states in AI technologies. The initiative, which brings together 14 countries and the European Union, will help participants establish practical cooperation and formulate common approaches to the development and implementation of AI. At the same time, it is a symptom of the growing technological rivalry in the world, primarily between the United States and China. Russia’s ability to interact with the GPAI may be limited for political reasons, but, from a practical point of view, cooperation would help the country implement its national AI strategy.

AI Brothers

The Global Partnership on Artificial Intelligence (GPAI) was officially launched on June 15, 2020, at the initiative of the G7 countries alongside Australia, India, Mexico, New Zealand, South Korea, Singapore, Slovenia and the European Union. According to the Joint Statement from the Founding Members, the GPAI is an “international and multistakeholder initiative to guide the responsible development and use of AI, grounded in human rights, inclusion, diversity, innovation, and economic growth.”

In order to achieve this goal, GPAI members will look to bridge the gap between theory and practice by supporting both research and applied activities in AI. Cooperation will take place in the form of working groups that will be made up of leading experts from industry, civil society and the public and private sectors and will also involve international organizations. There will be four working groups in total, with each group focusing on a specific AI issue: responsible AI; data governance; the future of work; and innovation and commercialization. In acknowledgment of the current situation around the world, the partners also included the issue of using AI to overcome the socioeconomic effects of the novel coronavirus pandemic in the GPAI agenda.

In terms of organization, the GPAI’s work will be supported by a Secretariat to be hosted by the Organisation for Economic Co-Operation and Development (OECD) and Centres of Expertise – one each in Montreal and Paris.

To better understand how this structure came to be, it is useful to look at the history of the GPAI itself. The idea was first put forward by France and Canada in June 2018, when, on the eve of the G7 Summit, Justin Trudeau and Emmanuel Macron announced the signing of the Canada–France Statement on Artificial Intelligence, which called for the creation of an international group to study AI-related issues. By that time, both countries had already adopted their own national AI development strategies – Canada was actually the first country in the world to do so in March 2017. The two countries proposed a mandate for the international group, then known as the International Panel on Artificial Intelligence, at the G7 conference on artificial intelligence in late 2018. A declaration on the creation of the group was then made in May 2019, following a meeting of the G7 Ministers responsible for digital issues. The group was expected to be formally launched three months later at the G7 Summit in Biarritz, with other interested countries (such as India and New Zealand) joining.

However, the initiative did not receive the support of the United States wfithin the G7. Donald Trump and Emmanuel Macron were expected to announce the launch of the group at the end of the event, but the American delegation blocked the move. According to Lynne Parker, Deputy Chief Technology Officer at the White House, the United States is concerned that the group would slow down the development of AI technology and believes that it would duplicate the OECD’s work in the area. The originators of the idea to create the group (which received the name Global Partnership on Artificial Intelligence in Biarritz) clearly took this latter point into account, announcing that the initiative would be developed under the auspices of the OECD.

A Principled Partnership

Like other international structures, the OECD has started to pay greater attention to artificial intelligence in recent years, with its most important achievement in this area being the adoption of the Recommendation of the Council on Artificial Intelligence. Unlike other sets of principles on AI, the OECD’s recommendations were supported by the governments of all member countries, as well as by Argentina, Brazil, Colombia, Costa Rica, Peru and Romania, which made it the first international document of its kind. They were also used as the basis for the Global Partnership on Artificial Intelligence.

In accordance with the OECD recommendations, signatory countries will adhere to the following principles of AI development: promote AI technologies for inclusive growth, sustainable development and well-being; the priority of human-centred values and fairness throughout the life-cycle of AI systems; the transparency and (maximum possible) explainability of AI algorithms; the robustness, security and safety of AI systems; and the accountability of AI actors.

In addition to this, the document proposes that the following factors be taken into account when drafting national AI development strategies: investing in AI research and development; fostering a digital ecosystem for AI research and the practical implementation of AI technologies (including the necessary infrastructure); shaping national policies that allow for a smooth transition from theory to practice; building human capacity and preparing for labour market transformation; and expanding international cooperation in AI.

A few weeks after the OECD endorsement, the recommendations on AI were included as an annex to the G20 Ministerial Statement on Trade and Digital Economy dated July 9, 2019, albeit with slightly different wording. The principles thus received the support of Russia, China and India.

Within the OECD itself, the recommendations served as an impetus for the creation of the OECD AI Policy Observatory (OECD.AI), a platform for collecting and analysing information about AI and building dialogue with governments and other stakeholders. The platform will also be used within the framework of the Global Partnership on Artificial Intelligence.

Artificial Intelligence and Realpolitik

The decision of the United States to join the GPAI was likely motivated more by political reasons than anything else. In the run-up of the G7 Science and Technology Ministers’ Meeting in late May 2020 (where all participants, including the United States, officially announced the launch of the GPAI), Chief Technology Officer of the United States Michael Kratsios published an article in which he stated that democratic countries should unite in the development of AI on the basis of fundamental rights and shared values, rather than abuse AI to control their populations, which is what authoritarian regimes such as China do. According to Kratsios, it is democratic principles that unite the members of the GPAI. At the same time, Kratsios argues that the new coalition will not be a standard-setting or policy-making body, that is, it will not be a regulator in the field of AI.

The United States Strategic Approach to the People’s Republic of China published in May 2020 and the many practical steps that the American side has taken in recent years are a reflection of the tech war currently being waged between the United States and China. For example, the United States has taken a similar approach to the formation of new coalitions in the context of 5G technologies. In 2018–2019, the United States actively pushed the narrative that the solutions offered by Huawei for the creation of fifth-generation communications networks were not secure and convinced its allies to not work with Beijing. Thirty-two countries supported the recommendations put forward at the Prague 5G Security Conference in May 2019 (the Prague Proposals), which included ideas spread by the United States during its campaign against Huawei (for example, concerns about third countries influencing equipment suppliers).

The United States is not the only GPAI member that is concerned about China. Speaking back in January about the U.S. doubts regarding the Franco–Canadian initiative, Minister for Digital Affairs of France Cédric O noted, “If you don’t want a Chinese model in western countries, for instance, to use AI to control your population, then you need to set up some rules that must be common.” India’s participation in the GPAI is particularly telling, as the United States has been trying to involve India in containing China in recent years. The new association has brought together all the participants in the Quadrilateral Security Dialogue (Australia, India, the United States and Japan), which has always been a source of concern for Beijing, thus sending a very clear signal to the Chinese leadership.

The Prospects for Russia

The political logic that guides the United States when it comes to participating in the Global Partnership on Artificial Intelligence may very well extend to Russia. The Trump administration formally declared the return of great power competition in its 2017 National Security Strategy. In Washington, Russia and China are often referred to as the main rivals of the United States, promoting anti-American values.

When assessing the possibility of interaction between Russia and the GPAI, we need to look further than the political positions of the participants. According to the Joint Statement from the Founding Members, the GPAI is open to working with other interested countries and partners. In this regard, the obvious points of intersection between Russia and the new association may produce favourable conditions for practical cooperation in the future.

First of all, the GPAI members and Moscow rely on the same principles of AI development. Russia indirectly adopted the OECD recommendations on artificial intelligence when it approved the inclusion of the majority of their provisions in the Annex to the G20 Ministerial Statement on Trade and Digital Economy in 2019 and thus shares a common intention to ensure the responsible and human-centred development and use of artificial intelligence technologies. This does not mean that there will not be differences of opinion of specific issues, but, as we have already noted, in its current form, the activities of the GPAI will not be aimed at unifying the approaches of the participants.

Second, according to media reports, Russia is working to re-establish ties with the OECD. It is already helping the OECD with its website, periodically providing data on new legal documents that will create a framework for the development and implementation of AI that have been adopted or are being considered.

Third, the current development of the national AI ecosystem in Russia shows that the state, business and the scientific community are interested in the same topics that are on GPAI agenda. This is reflected in the National Strategy for the Development of Artificial Intelligence for the Period up to the Year 2030 adopted in October 2019 and the draft Federal Project on the Development of Artificial Intelligence as Part of the National Programme “Digital Economy of the Russian Federation.” Furthermore, following the adoption of the National Strategy last year, Russian tech companies set up an alliance for AI development in conjunction with the Russian Direct Investment Fund, which is very much in keeping with the multistakeholder approach adopted by the Global Partnership on Artificial Intelligence.

It would seem that politics is the main stumbling block when it comes to Russia’s possible participation in GPAI initiatives, for example, the organization’s clear anti-Chinese leaning or its members openly discrediting Russia’s approaches to the development of AI. That said, Russia has nothing to gain from politicizing the GPAI, since cooperation with the organization could help it achieve its own goals in artificial intelligence. What is more, we cannot rule out the possibility that the GPAI will be responsible in the future for developing unified AI rules and standards. It is in Russia’s interests to have its voice heard in this process to ensure that these standards do not turn into yet another dividing line.

*Evgeniya Drozhashchikh, Ph.D. Student in the Faculty of World Politics at Lomonosov Moscow State University, RIAC Expert

From our partner RIAC

Continue Reading
Comments

Science & Technology

What is a ‘vaccine passport’ and will you need one the next time you travel?

Published

on

An Arab-Israeli woman shows her COVID-19 card which shows she has been vaccinated against the virus. Mohamed Yassin

Is the idea of a vaccine passport entirely new?

The concept of a passport to allow for cross border travel is something that we’ve been working on with the Common Trust Network for many months. The focus has been first on diagnostics. That’s where we worked with an organization called “The Commons Project” to develop the “Common Trust Framework”. This is a set of registries of trusted data sources, a registry of labs accredited to run tests and a registry of up-to-date border crossing regulations.

The set of registries can be used to generate certificates of compliance to prevailing border-crossing regulations as defined by governments. There are different tools to generate the certificates, and the diversity of their authentication solutions and the way they protect data privacy is quite remarkable.

We at the Forum have no preference when it comes to who is running the certification algorithm, we simply want to promote a unique set of registries to avoid unnecessary replication efforts. This is where we support the Common Trust Framework. For instance, the Common Pass is one authentication solution – but there are others, for example developed by Abbott, AOK, SICPA (Certus), IBM and others.

How does the system work and how could it be applied to vaccines?

The Common Trust Network, supported by the Forum, is combining the set of registries that are going to enrol all participating labs. Separately from that, it provides an up-to-date database of all prevailing border entry rules (which fluctuate and differ from country to country).

Combining these two datasets provides a QR code that border entry authorities can trust. It doesn’t reveal any personal health data – it tells you about compliance of results versus border entry requirements for a particular country. So, if your border control rules say that you need to take a test of a certain nature within 72 hours prior to arrival, the tool will confirm whether the traveller has taken that corresponding test in a trusted laboratory, and the test was indeed performed less than three days prior to landing.

The purpose is to create a common good that many authentication providers can use and to provide anyone, in a very agnostic fashion, with access to those registries.

What is the WHO’s role?

There is currently an effort at the WHO to create standards that would process data on the types of vaccinations, how these are channelled into health and healthcare systems registries, the use cases – beyond the management of vaccination campaigns – include border control but also possibly in the future access to stadia or large events. By establishing in a truly ethical fashion harmonized standards, we can avoid a scenario whereby you create two classes of citizens – those who have been vaccinated and those who have not.

So rather than building a set of rules that would be left to the interpretation of member states or private-sector operators like cruises, airlines or conveners of gatherings, we support the WHO’s effort to create a standard for member states for requesting vaccinations and how it would permit the various kinds of use cases.

It is important that we rely on the normative body (the WHO) to create the vaccine credential requirements. The Forum is involved in the WHO taskforce to reflect on those standards and think about how they would be used. The WHO’s goal is to deploy standards and recommendations by mid-March 2021, and the hope is that they will be more harmonized between member states than they have been to date in the field of diagnostics.

What about the private sector and separate initiatives?

When registry frameworks are being developed for authentication tools providers, they should at a minimum feed as experiments into the standardization efforts being driven by WHO, knowing that the final guidance from the only normative body with an official UN mandate may in turn force those providers to revise their own frameworks. We certainly support this type of interaction, as public- and private-sector collaboration is key to overcoming the global challenge posed by COVID-19.

What more needs to be done to ensure equitable distribution of vaccines?

As the WHO has warned, vaccine nationalism – or a hoarding and “me-first” approach to vaccine deployment – risks leaving “the world’s poorest and most vulnerable at risk.”

COVAX, supported by the World Economic Forum, is coordinated by the World Health Organization in partnership with GAVI, the Vaccine Alliance; CEPI, the Centre for Epidemics Preparedness Innovations and others. So far, 190 economies have signed up.

The Access to COVID-19 Tools Accelerator (ACT-Accelerator) is another partnership, with universal access and equity at its core, that has been successfully promoting global collaboration to accelerate the development, production and equitable access to COVID-19 tests, treatments and vaccines. The World Economic Forum is a member of the ACT-Accelerator’s Facilitation Council (governing body).

World Economic Forum

Continue Reading

Science & Technology

Iran among five pioneers of nanotechnology

Published

on

Prioritizing nanotechnology in Iran has led to this country’s steady placement among the five pioneers of the nanotechnology field in recent years, and approximately 20 percent of all articles provided by Iranian researchers in 2020 are relative to this area of technology.

Iran has been introduced as the 4th leading country in the world in the field of nanotechnology, publishing 11,546 scientific articles in 2020.

The country held a 6 percent share of the world’s total nanotechnology articles, according to StatNano’s monthly evaluation accomplished in WoS databases.

There are 227 companies in Iran registered in the WoS databases, manufacturing 419 products, mainly in the fields of construction, textile, medicine, home appliances, automotive, and food.

According to the data, 31 Iranian universities and research centers published more than 50 nano-articles in the last year. 

In line with China’s trend in the past few years, this country is placed in the first stage with 78,000 nano-articles (more than 40 percent of all nano-articles in 2020), and the U.S. is at the next stage with 24,425 papers. These countries have published nearly half of the whole world’s nano-articles.

In the following, India with 9 percent, Iran with 6 percent, and South Korea and Germany with 5 percent are the other head publishers, respectively.

Almost 9 percent of the whole scientific publications of 2020, indexed in the Web of Science database, have been relevant to nanotechnology.

There have been 191,304 nano-articles indexed in WoS that had to have a 9 percent growth compared to last year. The mentioned articles are 8.8 percent of the whole produced papers in 2020.

Iran ranked 43rd among the 100 most vibrant clusters of science and technology (S&T) worldwide for the third consecutive year, according to the Global Innovation Index (GII) 2020 report.

The country experienced a three-level improvement compared to 2019.

Iran’s share of the world’s top scientific articles is 3 percent, Gholam Hossein Rahimi She’erbaf, the deputy science minister, has announced.

The country’s share in the whole publications worldwide is 2 percent, he noted, highlighting, for the first three consecutive years, Iran has been ranked first in terms of quantity and quality of articles among Islamic countries.

Sourena Sattari, vice president for science and technology has said that Iran is playing the leading role in the region in the fields of fintech, ICT, stem cell, aerospace, and is unrivaled in artificial intelligence.

From our partner Tehran Times

Continue Reading

Science & Technology

Free And Equal Internet Access As A Human Right

Published

on

Having internet access in a free and equal way is very important in contemporary world. Today, there are more than 4 billion people who are using internet all around the world. Internet has become a very important medium by which the right to freedom of speech and the right to reach information can be exercised. Internet has a central tool in commerce, education and culture.

Providing solutions to develop effective policies for both internet safety and equal Internet access must be the first priority of governments. The Internet offers individuals power to seek and impart information thus states and organizations like UN have important roles in promoting and protecting Internet safety. States and international organizations play a key role to ensure free and equal Internet access.

The concept of “network neutrality is significant while analyzing equal access to Internet and state policies regulating it. Network Neutrality (NN) can be defined as the rule meaning all electronic communications and platforms should be exercised in a non-discriminatory way regardless of their type, content or origin. The importance of NN has been evident in COVID-19 pandemic when millions of students in underdeveloped regions got victimized due to the lack of access to online education.

 Article 19/2 of the International Covenant on Civil and Political Rights notes the following:

“Everyone shall have the right to freedom of expression; this right shall include freedom to seek, receive and impart information and ideas of all kinds, regardless of frontiers either orally, in writing or in print, in the form of art, or through any other media of his choice.”

Internet access and network neutrality directly affect human rights. The lack of NN undermines human rights and causes basic human right violations like violating freedom of speech and freedom to reach information. There must be effective policies to pursue NN. Both nation-states and international organizations have important roles in making Internet free, safe and equally reachable for the people worldwide. States should take steps for promoting equal opportunities, including gender equality, in the design and implementation of information and technology. The governments should create and maintain, in law and in practice, a safe and enabling online environment in accordance with human rights.

It is known that, the whole world has a reliance on internet that makes it easy to fullfill basic civil tasks but this is also threatened by increasing personal and societal cyber security threats. In this regard, states must fulfill their commitment to develop effective policies to attain universal access to the Internet in a safe way.

 As final remarks, it can be said that, Internet access should be free and equal for everyone. Creating effective tools to attain universal access to the Internet cannot be done only by states themselves. Actors like UN and EU have a major role in this process as well.

Continue Reading

Publications

Latest

Trending