THE IMPACT OF ARTIFICIAL INTELLIGENCE AND GENERATIVE NEURAL NETWORKS ON MODERN WORLD POLITICS

ВЛИЯНИЕ ИСКУССТВЕННОГО ИНТЕЛЛЕКТА И ГЕНЕРАТИВНЫХ НЕЙРОННЫХ СЕТЕЙ НА СОВРЕМЕННУЮ МИРОВУЮ ПОЛИТИКУ

Эзе Бонавентура Огона

студент, кафедра международных отдела, Нигерийский университет,

Федеративная Республика Нигерия, г. Нсукка

ABSTRACT

The article proposes a general overview of the impact of artificial intelligence and generative neural systems on modern world politics in the context of events in the world of politics over the past 5 years. The article touches upon such areas as the impact of Artificial Intelligence and generative neural systems technologies on the election process and its risks, and also describes the main competitors of the United States and China in the development of Artificial Intelligence and generative neural systems technologies, their interaction features, the main stages of technology development in these countries, as well as the features of the focus of the United States and China in the development of Artificial Intelligence and generative neural systems technologies depending on their political systems.

АННОТАЦИЯ

В статье предлагается общий обзор влияния искусственного интеллекта и генеративных нейронных систем на современную мировую политику в контексте событий в мире политики за последние 5 лет. Статья затрагивает такие сферы как влияние Искусственного интеллекта и генеративных нейронных систем технологий на процесс выборов и его риски, а также описывает основных конкурентов США и Китая, в развитии технологий Искусственного интеллекта и генеративных нейронных систем, их особенности взаимодействия, основные этапы развития технологий в этих странах, а также особенности фокуса внимания США и Китая в технологиях развития Искусственного интеллекта и генеративных нейронных систем в зависимости от их политических систем.

Keywords: artificial intelligence (ai), geopolitics, global policy, national policy, misinformation, disinformation, fact-checking.

Ключевые слова: искусственный интеллект (ии), геополитика, глобальная политика, национальная политика, дезинформация, проверка фактов.

Introduction

Over the past couple of years, the global information field has been under intense influence of Artificial Intelligence (AI) technologies. Due to the high variability of the ways of using AI and its great potential, both domestic and foreign policies of countries have been influenced by it. As a result, the use of AI in local and global politics threatens the democratic discourse of international relations, as well as the electoral order. In view of the above observations, the topic of AI implementation requires further study and discussion.

The object of this study is the role and prospects of AI in political processes at the global and local levels. The subject of the study is the advantages and disadvantages of current methods of implementing and using AI in the political sphere at the present stage.

The aim of the study is to examine examples of the use of AI and its role in modern political and electoral processes at the local and global levels and to predict subsequent developments taking into account the current facts of the use of AI in modern politics.

To achieve the stated goal, the following research objectives have been formulated:

1. To study the risk of unlimited use of AI in the political sphere at the national level;
2. To analyze the risk of unlimited use of AI for the sphere of politics at the global level;
3. To study the main global competitors in the development of AI technologies and their achievements;
4. To propose options for the development of events in the current situation of introducing AI into political processes at the international and local levels.

The Raise of Artificial Intelligence

Artificial intelligence, according to the definition of P. Russell and P. Norving, is the ability of abstract thinking, logic, self-awareness among machines, especially among electronic computing machines. This is a large field study included in computer science. AI includes the study of such issues as methods and software applicable to machines in order to improve their ability to study, understand the environment and self-learn [20, p. 3].

Artificial intelligence has been constantly evolving towards its current level of autonomy. Until the 21st century, AI needed the support of human intelligence to perform tasks, but the 21st century was a dividing line after which AI began to exhibit generative skills, AI systems such as OpenAI's ChatGPT. In fact, AI and generative neural systems have moved from a theoretical form to a practical one. This, in turn, has directly influenced political discourse and decision-making processes [17]. This can be seen from the example of the changed geopolitical and social picture of the world in the last three years. More detailed examples confirming this statement will be presented in the following subheading.

Artificial Intelligence and its Impact on Domestic Politics and the Social Sphere

The public and political spheres were objectively unprepared for the technological leap in the development of AI. Some plans and expectations for the results of using AI and generative neural systems did not come true. Given that 2024 has become a record year for the number of countries in which elections will be held, multiple facts about the influence of AI on the course of elections are already presented by real examples [9].

One of them is a deepfake of President Biden calling on New Hampshire Democrats not to vote earlier this year [7]. The author of the deepfake has not been found, but one of the reasons for what happened can be considered the wide availability of AI programs and software for ordinary people. Some of these programs and applications are freely available or require only a small fee for full access to the AI toolkit.

The very first example of AI usage in an electoral campaign in a country like the United State of America was during the president Barack Obama's campaign in 2012 [14]. AI was used to calculate the best day, select the best state, and calculate the audience for presidential candidate Barack Obama's public appearance. According to the study, this increased the number of votes by 10-12%, which allowed B. Obama to break away from his competitors [13].

Illegal collection of electorate data has also become a common use of AI tools in election campaigns since 2016. For example, in 2016, Cambridge Analytica, a company that closely cooperates with Trump's team, managed to collect data from 87 million Facebook users (banned in the Russian Federation). This allowed them to analyze the audience, which also influenced the election results [5].

Also, according to the results of a 2020 Oxford University study, 81 countries used neural-generative technologies to manipulate public opinion in political matters [4, p.11]. The study mentions the use of the following AI tools: copying Face ID databases, microtargeting, copying user votes, chatbots, and using content generation algorithms. If the previous examples of AI risks concerned the election campaign, then there is one that is directly related to the voting process. In addition to the existing possibility of creating synthetic content, there is a threat of manipulating real content through microtargeting of certain groups of the population. Platforms such as Meta (banned in the Russian Federation) and Google*(At the request of Roskomnadzor, we inform you that the foreign entity that owns Google information resources is a violator of the legislation of the Russian Federation – editor’s note.) have been using the model they created for collecting and processing user data for several years. This is done with the aim of creating personalized selections of content and advertising for each user [18, p. 10]. At the moment, the platforms' algorithm is as follows: hypothesis testing and randomized user selection. Algorithms offer content and track the user's level of interest in it in order to subsequently exclude it or offer similar content [11]. However, with the advent of AI, the collected information can be used radically for other purposes, limiting the democratic rights of certain groups of the population, national minorities, etc. The customer of propaganda will have the opportunity to tailor the content to certain micro-groups, a specific race / religion / sports preferences and even age [12]. For example, for your purposes, you can generate content for male representatives with brown skin, 25 years old, living in New York state and supporting the Atlanta Hawks basketball team [21].

Thus, in view of the above risks, AI can become a threat to the democratic system of the state. This is due to the general availability of programs running on AI, the lack of legislative regulation in the field of AI technologies, as well as the low level of media and information literacy among the population.

Geopolitics in the world using generative AI. Main rivals and their challenges

In the field of global politics, AI has also made significant adjustments and is changing the world order. According to Stanford University’s 2023 Artificial Intelligence Index Report, the United States and China, in addition to geopolitical confrontation, are the main competitors in the development of AI technologies. According to this report, cooperation between the United States and China in the development of AI technology from 2010 to 2021 increased fourfold, but crossing the threshold of these years, the interaction sharply decreased. The report predicts a subsequent decline in cooperation in the field of AI [2, p. 23]. Competition in AI development affects technological, economic, military and political spheres.

Speaking in more detail about the subject of competition, it is worth mentioning the main field of AI technology development in which China and the United States compete - Large language models (LLMs). The most important criterion for assessing the development of LLMs is the number of parameters - the larger the number of parameters, the more practical information, therefore more calculations and more accurate the result. By the way, ChatGPT4 uses up to 1 trillion parameters [1]. By this criterion, China is slightly behind the US. The US is traditionally the birthplace of innovators in this area, and world-famous IT companies and valleys are also located in the US [6].

The second significant parameter for the success of AI technology development is hardware supplements. This is especially related to the availability of semiconductors or chips to create a powerful connection and access to graphics processing units (GPUs). The confrontation between China and the United States in recent years has been quite strongly focused on hardware. While China has traditionally developed a semiconductor industry and fully satisfies the needs of production, the production of microchips, without which it is impossible to launch AI systems, is a vulnerability of the AI sphere for China [3]. Thus, the US found leverage in the form of China's technological shortage, and in 2022 a series of measures followed on the part of the US on the export of microchips, in which the country cooperated with Japan and the Netherlands. This is in line with the "small yard, high fence" strategy, which stands for the protection of technologies sensitive to national security [19].

By 2023, the United States again announced a series of measures to control technologies sensitive to national security, which again affected the production of semiconductors and microchips. In turn, China responded with counter-sanctions on the export of inert metals gallium, germanium and graphite - raw materials for the production of semiconductors and other hardware components. More importantly, after these sanctions, China turned its attention to the development of its technology production, increased the number of foreign orders for the production of semiconductor raw materials [22, p. 41].

Thus, despite sufficient mutual sanctioning, China continues to move towards improving AI technology in an autonomous mode, although at a slower pace than the United States.

Finally, development in the field of AI, as a new global force of influence, also depends on the traditional domestic political structure of the country. The example of China shows that the political system of this country is quite conservative. If the United States is an example of a country with an open type of society, then the country's challenges to AI will be corresponding. For the United States, the main challenges in the development of AI technologies today are the imperfections of the LLMs system or so-called hallucinations (failures and erroneous results of the system). China is an example of a country with a closed type of society. Here, the main challenges are the ability to control AI content that is not approved by the authorities, consumed and generated by the population. It is also important for Beijing to retain the ability to promote its policy in the context of AI development [23]. Thus, in 2023, the PRC established a procedure for using AI for developers and users called Measures on the Administration of Generative Artificial Intelligence Services (Draft for Solicitation of Comments). It states that “Content generated using generative AI should reflect the core socialist values and must not contain content that subverts the state power, overthrows the socialist system, incites secession, undermines national unity, promotes terrorism, extremism, promotes ethnic hatred, ethnic discrimination, violence, obscene and pornographic information, false information, or content that may disrupt economic and social order” [16].

Conclusion

In the last couple of years, there has been an accelerated leap in the development of AI technologies and generative neural systems. AI has become an integral part of social, economic and political life. This is clearly seen in the example of past years of elections in the United States and related high-profile cases. AI intervention in politics is manifested in the form of interference in legislative election processes, targeting micro groups of the population and influencing their voting. The phenomenon of deepfakes is also observed everywhere. The increase in the level of misinformation and disinformation through the use of AI will lead to global social and political consequences, such as inciting ethnic hatred, increasing the number of anti-vaccination followers, misleading the campaign of presidential candidates and much more. It follows that at the moment there is an urgent need to unite the world community to develop norms and laws regulating and controlling the use of AI for the generation and consumption of content. Technology companies are also responsible. The main rivals in the development of AI technologies are the United States and China. However, these two countries are also the main comrades in developing ways to control the use of AI. Being a new force, akin to nuclear weapons, AI launches a kind of "Arms Race", or in modern language, a Digital Divide, in which a competitor with a more powerful set of AI tools will be able to control the geopolitical situation, the direction of sanctions, economic flows, trade, etc.

References:

1. Albergotti R., The secret history of Elon Musk, Sam Altman, and OpenAI, Semafor, 2024 https://www.semafor.com/article/03/24/2023/the-secret-history-of-elon-musk-sam-altman-and-openai
2. Artificial Intelligence Index Report 2023, Stanford Unversity https://aiindex.stanford.edu/wp-content/uploads/2023/04/HAI_AI-Index-Report_2023.pdf
3. Atkinson R. China Is Rapidly Becoming a Leading Innovator in Advanced Industries, Information Technology & Innovation Foundation, 2024 https://itif.org/publications/2024/09/16/china-is-rapidly-becoming-a-leading-innovator-in-advanced-industries/
4. Bradshaw S., Bailey H., Howard F., Industrialized Disinformation 2020 Global Inventory of Organized Social Media Manipulation\ Oxford, 2020 https://demtech.oii.ox.ac.uk/wp-content/uploads/sites/12/2021/02/CyberTroop-Report20-Draft9.pdf
5. Capraro V., Lentsch A., The impact of generative artificial intelligence on socioeconomic inequalities and policy making, 2024 https://www.researchgate.net/publication/381331889_The_impact_of_generative_artificial_intelligence_on_socioeconomic_inequalities_and_policy_making
6. Cohen J., Lee G., The generative world order: AI, geopolitics, and power, Goldman Sachs, 2023 https://www.goldmansachs.com/insights/articles/the-generative-world-order-ai-geopolitics-and-power
7. Collier K., Wong S.., Fake Biden robocall telling Democrats not to vote is likely an AI-generated deepfake, NBC News https://www.nbcnews.com/tech/misinformation/joe-biden-new-hampshire-robocall-fake-voice-deep-ai-primary-rcna135120
8. Crenshaw K., Demarginalizing the Intersection of Race and Sex: A Black Feminist Critique of Antidiscrimination Doctrine, Feminist Theory and Antiracist Politics, 1989 U. CHI. LEGAL F. 139.
9. Curran E., Crawford A., Brace for Elections: 40 Countries Are Voting in 2024, Bloomberg https://www.bloomberg.com/news/articles/2023-11-01/brace-for-elections-40-countries-are-voting-in-2024
10. FOUNDATION (Jan. 27, 2022), https://il.boell.org/en/2022/01/27/ai-and-elections-observationsanalyses-and-prospects
11. Goldstein, supra note 54, at 24
12. Goldstein, supra note 54, at 63
13. Issenberg S., How Obama’s Team Used Big Data to Rally, 2012 Votershttps://www.technologyreview.com/2012/12/19/114510/how-obamas-team-used-big-data-to-rally-voters/
14. Jack Virag, A/B Testing and Experimentation in the Obama 2008 and 2012 Campaigns, STATSIG (Oct 25. 2022), https://www.statsig.com/blog/data-experimentation-testing-obama-electioncampaigns); Arunesh Mathur et al.,
15. Manipulative Tactics Are the Norm in Political Emails: Evidence from 300K Emails from the 2020 US Election Cycle, BIG DATA & SOC. 1 (2023).
16. Measures on the Administration of Generative Artificial Intelligence Services (Draft for Solicitation of Comments), China Internet Information Office, 2023 https://www.cac.gov.cn/2023-04/11/c_1682854275475410.htm?_x_tr_sl=auto&_x_tr_tl=ru&_x_tr_pto=wapp&_x_tr_hl=uk
17. Namitha A. 2023 Generative AI in the Government Sector: Use-cases, Challenges & Best Practices https://www.rezolve.ai/blog/generative-ai-government#top-10-use-cases-of-generative-ai-in-the-government
18. Overton S., Overcoming Racial Harms to Democracy from Artificial Intelligence, GW Law, 2024 https://scholarship.law.gwu.edu/cgi/viewcontent.cgi?article=2986&context=faculty_publications
19. Remarks by National Security Advisor Jake Sullivan on the Biden-⁠Harris Administration’s National Security Strategy, The White House, 2022 https://www.whitehouse.gov/briefing-room/speeches-remarks/2022/10/13/remarks-by-national-security-advisor-jake-sullivan-on-the-biden-harris-administrations-national-security-strategy/
20. Russell, Stuart J.; Norvig, Peter (2021). Artificial Intelligence: A Modern Approach (4th ed.). Hoboken: Pearson. ISBN 978-0-1346-1099-3. LCCN 20190474. https://dl.ebooksworld.ir/books/Artificial.Intelligence.A.Modern.Approach.4th.Edition.Peter.Norvig.%20Stuart.Russell.Pearson.9780134610993.EBooksWorld.ir.pdf
21. Sophia Cai, How AI is Already Changing the 2024 Election, AXIOS (Apr. 29, 2023), https://www.axios.com/2023/04/29/how-ai-already-changing-2024-election
22. The US-China Economic and Security Review Commission, One Hundred Eighteenth Congress, 2023 https://www.uscc.gov/sites/default/files/2023-11/2023_Annual_Report_to_Congress.pdf
23. Thibaut K., Chinese discourse power: Ambitions and reality in the digital domain, In-Depth Research & Reports, 2022 https://www.atlanticcouncil.org/in-depth-research-reports/report/chinese-discourse-power-ambitions-and-reality-in-the-digital-domain/