SummaryAs significant actors influencing geopolitical dynamics, the role of technology companies has not received adequate attention in academia. As technology increasingly becomes a strategic tool for sovereign states to maintain interests and exert influence in competition, the power dynamics and interdependencies between technology companies and sovereign states are undergoing profound changes. Leveraging their leading role in technological transformation, technology companies are becoming key variables affecting national innovation capacity, domestic governance capabilities, and comprehensive national power. In this process, technology companies are no longer merely passive responders; instead, they are proactively adapting to new dynamics, influencing policy-making through active involvement in geopolitical struggles to achieve their commercial interests. Through in-depth interviews with 25 practitioners in the field of artificial intelligence, combined with case studies of specific actions taken by Intel, Mistral AI, and OpenAI, this research reveals the behavioral logic of technology companies’ proactive adaptation and the potential risks arising from resource allocation imbalances, poor regulatory effectiveness, damaged international cooperation, and distorted ecosystems. Therefore, states should cautiously respond to the power and interest conflicts brought about by the rise of tech giants while emphasizing the pivotal role of technology companies in shaping the global political and economic landscape, fostering international cooperation with an open and inclusive attitude, and jointly guiding technological innovation for the benefit of humanity.
KeywordsTechnology Companies, Artificial Intelligence, Great Power Competition, Technology Regulation, Global Governance
With the rapid development and widespread application of cutting-edge technologies such as artificial intelligence, technology companies have become not only drivers of economic development but also critical players in the international political landscape. The breakthroughs in generative artificial intelligence and its immense potential further highlight the influence of technology companies on economic, cultural, and social environments. In the context of great power competition, the key influence of technology companies on the strategic games among major powers has garnered widespread attention from various countries. The power dynamics and interdependencies between technology companies and sovereign states are undergoing profound changes.
Currently, research on technology companies in the context of great power competition primarily focuses on three aspects: first, exploring the impact of technology on the power and status of technology companies, such as data control and platform characteristics; second, examining the influence of technological advancement on the relationship between technology companies and states, which includes both cooperation and conflict; third, analyzing the actions of technology companies under national competition, with much of the research focusing on how technology companies passively respond to geopolitical changes and national policy pressures. Overall, current studies tend to view technology companies as passive recipients and strategic tools in great power competition, while there is relatively insufficient attention and analysis on their active intervention in geopolitical affairs as independent actors.
However, since the outbreak of the Ukraine crisis, scholars have noted that technology companies have actively intervened in geopolitical actions during conflicts, viewing this as a critical historical juncture in the role transformation of technology companies against the backdrop of changing international dynamics. In fact, amid intensified great power competition, while states may adjust market logic based on power logic to maintain comparative power advantages, technology companies may also actively exploit state logic to expand their commercial interests. In light of this, this article focuses on technology companies in the era of technological change, exploring how technology reshapes the dependency relationship between states and technology companies, thereby enhancing the influence of technology companies on national policies, and how technology companies proactively leverage geopolitical dynamics to influence policy-making under the context of great power competition, along with the potential risks associated with this trend.
To address these issues, this article will analyze the evolution of the relationship between technology companies and states in the context of great power competition based on the history of technology companies, and examine the behavioral logic of technology companies influencing policy-making through this dynamic. The case studies will focus on the actions and underlying commercial logic of Intel, Mistral AI, and OpenAI in influencing policies, further exploring the potential risks associated with such behaviors. To enhance the reliability of the analysis, this article will supplement official public materials and academic literature with in-depth interviews with practitioners in the field of artificial intelligence.
1. The Restructuring of the Relationship Between Technology Companies and States in the Context of Great Power Competition
By virtue of their leading role in technological transformation, technology companies are becoming key variables influencing national innovation capacity, governance capabilities, and comprehensive national power. In this process, the relationship between technology companies and states is undergoing profound changes, allowing technology companies to leverage the competitive landscape between states to influence policies.
(1) The Development History of Technology Companies
The development of large technology companies is closely linked to the trajectory of disruptive technologies, which have established the key influence of large technology companies in the current geopolitical landscape. According to scholars such as Carlota Perez and Reijer Hendrikse, the development history of large technology companies can be roughly divided into three stages, each characterized by its own features and overlapping to some extent. During the development process, the technological innovation of large technology companies has evolved from focusing on hardware and software development to centering on data processing and infrastructure, with the current emphasis generally on the development and application of artificial intelligence technologies as a strategic focus.
In the 1970s, many large technology companies concentrated their innovations primarily on hardware. The technologies of companies like IBM and Intel laid the groundwork for the development of personal computers, particularly Intel’s development of microprocessors heralding a new technological era. Companies like Apple and Microsoft, founded around the same time, shifted their focus to enhancing the adaptability and usability of computers, accelerating the onset of the software revolution. From the mid-1980s to the late 1990s, the software industry experienced unprecedented prosperity, exemplified by Microsoft’s launch of the Windows operating system, which solidified the company’s dominant position in the field.
In the 1990s, the development of the internet had a profound impact on the structure of the technology sector. Companies like Amazon and Google capitalized on the continuous evolution of digital space to create new business models, marking the beginning of the digital platform era. Entering the 21st century, the potential of the internet was further explored, with large technology companies gradually transforming into leaders in constructing comprehensive digital ecosystems. Notably, the release of the iPhone in 2007 not only ushered in a new era of mobile internet and app economy but also provided the data foundation for the arrival of the artificial intelligence era by expanding the scale and scope of data extraction. Around these “infrastructural core” technology giants, numerous digital platforms emerged to serve consumers, further accelerating the digital transformation of other industries.
Since 2012, benefitting from advancements in machine learning and neural network technologies, technology companies have increased their investments in artificial intelligence research and development. Leveraging their extensive customer bases and distribution networks, companies like Apple, Alphabet, Amazon, Microsoft, IBM, and Facebook have strong advantages in developing and applying artificial intelligence technologies. In this process, artificial intelligence technology has gradually permeated various business areas of technology companies, including search engines and cloud computing. This has not only revolutionized the provision of products and services but has also raised a series of important ethical and regulatory issues regarding privacy, algorithmic bias, and the nature of future work. The launch of ChatGPT by OpenAI in November 2022 sparked a wave of competition among technology companies to invest in generative artificial intelligence research and development. Current business and policy dynamics indicate that generative artificial intelligence technology has become central to the strategic and competitive landscape of technology companies. Given the disruptive impact of this technology, the technology companies leading this innovation are undoubtedly the focal point of international competition.
(2) The Restructuring of the Relationship Between Technology Companies and States
As technologies that alter economic and social structures, advanced technologies such as artificial intelligence enable technology companies to “determine how states project economic and military power, shape future employment, and redefine social contracts,” thereby constructing the global environment on which great power competition relies. Against the backdrop of intertwined technological transformation and great power competition, the power relations between technology companies and states are being restructured, reflecting an asymmetrical interdependence across three dimensions.
First, technology companies are in a dominant position in technological development, creating an asymmetry in the dependency relationship between technology companies and sovereign states. Amid the acceleration of technological innovation, states’ proprietary control over the technological innovation process is gradually weakening and shifting to emerging non-state actors represented by technology companies. By collecting data, commodifying data, and centralizing talent and computing resources, technology companies have mastered key elements of technological innovation, establishing themselves as leaders in technological innovation. Simultaneously, technology companies leverage their technological advantages to create entirely new virtual spaces. In this space, technology companies can determine how individuals, groups, and institutions use their services, how information is exchanged, and how power is allocated in the network through code design, commercial practices, and private regulatory means. The complexity and diversity of these rules further reinforce the technological innovation capabilities of technology companies. Currently, technology companies are leading the direction of frontier technology development through forward-looking technological layouts and innovation resource investments, occupying first-mover advantages and dominant positions in global technology industry competition. Among the 19 technologies listed in the key and emerging technology lists formulated by the Trump and Biden administrations in the United States, 17 are led by the private sector.
Taking artificial intelligence technology as an example, in recent years, this technology has made breakthrough advances in fields such as computer vision, speech recognition, and natural language processing. From traditional methods to deep learning, from convolutional neural networks (CNN), recurrent neural networks (RNN) to new models like Transformers and Diffusion, the performance metrics of artificial intelligence continue to be refreshed. In the wave of rapid technological evolution, technology giants represented by Google and Microsoft are undoubtedly at the forefront of innovation. Thanks to substantial financial strength, a reserve of high-end talent, and a vast accumulation of data, these companies can continuously invest massive amounts of research and development funds to tackle core technologies. Since 2019, among the top five companies with the highest number of foundational models released, Google ranks first with 40 models, while OpenAI ranks second with 20 models. Foundational models like GPT-4, Claude 3, and Llama 2 represent the forefront of artificial intelligence and have been deployed in various real-world scenarios due to their powerful capabilities.
Unlike traditional bureaucratic structures, technology companies possess agility and adaptability, giving them unique advantages in driving key technological innovations. Through multinational cooperation and extensive layouts, technology companies effectively integrate global talent and resources. In the development of artificial intelligence, technology companies have fully leveraged thriving open-source communities, further accelerating the diffusion and iteration of technology. The development of artificial intelligence technology has a history of open-source collaboration, with foundational frameworks, datasets, and model codes being freely shared and continuously updated, enriched, and improved in online repositories like GitHub and Hugging Face. For example, Google’s TensorFlow and Facebook’s PyTorch are widely used in academia and industry and continue to evolve to meet the changing demands of machine learning and deep learning.
Thus, the innovation capabilities of technology companies are crucial for a nation’s position in the global technology competition landscape. In today’s world, technological innovation capability has become an essential component of national strength, closely linked to the positive shaping of national image.
Second, the irreplaceability of technology companies in supporting governments to provide public services and conduct governance work is rising, further strengthening states’ dependency on technology companies. As data becomes a critical element of national governance, cutting-edge technologies represented by artificial intelligence and big data analytics are reshaping how governments perceive social needs, optimize public resource allocation, and improve public service provision, enhancing their ability to meet citizens’ demands for public services. In this process, technology companies deeply engage in all aspects of government functions by providing technical support and data services, expanding their influence on public decision-making and the operation of public power. When formulating significant public policies, governments often require data support and model analysis from technology companies to better justify the necessity and feasibility of policies. Additionally, many public service matters also need to be realized through purchasing technology companies’ products and services.
It is worth noting that the influence gained by technology companies through technological means partially compensates for the regulatory gaps faced by states in the absence of digital governance rules. Large technology companies guide user and consumer behavior through various service terms, compliance statements, privacy policies, and data policies, effectively acquiring a form of “quasi-government” governance power. In the absence of relevant legislation, these mechanisms can play a practical role in regulating user behavior. As intelligent applications such as AI assistants and recommendation systems penetrate all aspects of social life, the rules constructed by companies through algorithms are having increasingly profound impacts on real society. In this trend, states need to leverage cooperation with technology companies to utilize their behavioral regulatory mechanisms to achieve public governance objectives.
At the same time, the impact of technology on ideology and social values, as well as its potential threats to national security and social order, have also attracted significant attention from policymakers. However, the rapid pace of technological development often poses challenges to states’ regulatory actions. This issue has been particularly prominent in the EU’s formulation of the “Artificial Intelligence Act.” The draft law proposed by the European Commission in April 2021 only considered the then-current level of technology, and with the release of models like ChatGPT, the EU was compelled to initiate a new round of negotiations to adapt to this situation. The European Commission’s High-Level Expert Group on Artificial Intelligence pointed out that due to the innovative nature of technology, insufficient understanding of its impacts, and unpredictable developments, there is currently a lack of evidence to guide relevant policy-making. In this context, the participation of technology companies at the forefront of technological development will help ensure that regulatory policies are based on the latest scientific research and professional insights, thereby better addressing the challenges posed by artificial intelligence technology. Thus, cooperation with technology companies is a crucial foundation for countries to formulate effective regulatory measures, while regulatory agencies also need to be vigilant against excessive reliance on the information and guidance provided by technology companies.
In summary, effective national governance is significantly influenced by the capabilities and willingness of technology companies. Currently, the dependency of states on technology companies in domestic governance is showing an increasing trend. Governments often need to collaborate with enterprises to achieve public governance goals and leverage the expertise and practical experience of enterprises for effective regulation.
Third, technology companies, as the backbone of driving innovation and industrial upgrading, are becoming crucial supports for national economic growth and enhancing international competitiveness. Many technological breakthroughs are often based on technological progress in a series of related industries, which in turn promotes the development of those industries. For instance, artificial intelligence is a highly integrative technology whose development relies on support from algorithms, computing power, data, and other aspects. Technology giants have accumulated vast data resources and possess rich datasets and analytical capabilities, providing necessary foundational support for training and optimizing artificial intelligence models. Additionally, since training artificial intelligence models requires substantial computing power, specialized chips for computing-intensive tasks like deep learning have become critical pillars of artificial intelligence technology, and the design and manufacturing technologies for such chips are also held by companies like NVIDIA and TSMC. With significant advancements in artificial intelligence technology, the development of semiconductor, cloud computing, and other technologies and related industries has gained enormous growth momentum.
Simultaneously, the technological breakthroughs and product innovations of technology companies provide new pathways and motivations for optimizing economic structures and transforming development methods. With the widespread application of artificial intelligence technology in fields such as finance, healthcare, and education, this technology is increasingly becoming a new engine for national economic growth and industrial transformation. It is estimated that generative artificial intelligence can contribute over $4 trillion to global economic growth each year. This potential impact indicates that artificial intelligence is not merely a technological issue but also a matter concerning macroeconomics and national competitiveness.
Furthermore, the level of technological innovation of technology companies is closely related to a country’s defense capabilities. Historically, countries that have dominated the international system are often those that excel at applying advanced technologies in military domains. Currently, artificial intelligence technology is widely applied in military equipment, intelligence analysis, and information command, and the driving force behind this technology’s innovation is represented by technology giants like Google, Amazon, and Meta, which possess unparalleled advantages in artificial intelligence research and development. Thus, the widespread application of emerging technologies in military domains further deepens states’ dependency on technology companies.
In conclusion, the cooperation between states and technology companies is not only a practical necessity for driving technological innovation and boosting national economies but also a strategic choice. The asymmetry of the dependency relationship between technology companies and states is concentrated in that while governments nominally occupy the center of sovereignty and governance, in practice, technology companies can leverage their leading positions in technological transformation and their ability to shape the competitive landscape between nations to exert significant influence on national policies and governance practices. In the context of significant changes in the power structure between technology companies and states, it is particularly necessary to reflect on the changes in the interactive relationships that arise.
2. Proactive Adaptation: The Behavioral Logic of Technology Companies Influencing National Policies
As commercial entities, the fundamental goal of technology companies influencing policies is to maximize profits. In the context of increasingly fierce global technological competition, the success of technology companies relies on continuous and sufficient resource support, a relatively lenient regulatory environment, and policy protection for their innovative outcomes, all of which often require governments to play a key role in relevant fields. Therefore, technology companies seek to influence policy-making to support their business expansion and commercial interests.
Existing research has discussed the influence of commercial actors on national policies extensively. The policy-making process of modern states is often influenced by various interest groups. As a special interest group, technology companies influence national decision-making primarily through lobbying, manifesting as “capturing” regulators to relax regulations and create favorable policy environments for their development. Although many technology giants traditionally adopt a position of technological liberalism, avoiding politics and lobbying, regulatory threats and incentives are prompting them to increasingly engage in policy-making. Statistics indicate that since 2014, among the three types of lobbyists (businesses, associations, and NGOs) contacting EU Commission staff, businesses have shown a significant upward trend, reaching 41% in 2018, with technology giants like Google, Microsoft, and Facebook being particularly prominent.
Clive S. Thomas defines lobbying as the interactive behavior aimed at benefiting specific groups or interests by influencing current policies through direct means (such as using lobbyists) or indirect means (such as public demonstrations). Technology companies employ various methods to influence policies, including participating in hearings, hiring experts closely cooperating with governments, and funding influential think tanks, universities, and experts. Essentially, policy-making is a communication process in which decision-makers communicate with social actors to advance goals and exchange underlying interests. In this process, technology companies need to consider how to identify specific dimensions of arguments and policy debates to influence decision-makers’ perceptions and promote policies that favor their interests. The “frame” is a core element of political communication and debate, defined as “selecting certain aspects of perceived reality and making them more prominent in communication texts.” Selecting a frame involves the process of choosing relevant important content, emphasizing specific interests and lobbying targets, and constructing arguments to achieve these goals. Actors utilize framing to promote decision-makers’ specific understandings of issues, thereby persuading them to integrate organizational goals into the final policies.
Existing research has analyzed the content of frames previously employed by technology companies. Relevant studies indicate that the power of technology companies’ platforms and the products and services they provide to governments and the public determine their political influence. Therefore, when lobbying for policies, technology companies tend to adopt a technological liberalism frame, emphasizing that technology can address many pressing social issues. However, such analyses focus on the interaction between technology companies and domestic politics, with insufficient discussion on technology companies’ involvement in foreign policy-making under the current international landscape. The intertwining of technological transformation and great power competition reshapes the relationship between technology companies and states, providing fertile ground for the expansion of technology companies’ political influence. This article will analyze how technology companies leverage the great power competition framework to influence policy-making based on the discussions of frames in the literature on policy lobbying.
According to Harold D. Lasswell’s policy cycle model, the policy steps include agenda setting, policy formulation, decision-making, policy implementation, and policy evaluation. Framing plays a critical role at each stage of the policy cycle, aiding in problem identification and solution formulation, and providing a basis for determining negotiation positions and setting standards for policy implementation and evaluation. This article primarily focuses on technology companies in emerging technology fields, where most relevant policies are still in exploratory stages. Therefore, the analysis of how the great power competition framework is used by technology companies to influence policy-making will mainly focus on three key links—problem definition, policy goal setting, and specific content design.
First, in the problem definition stage, technology companies can influence policymakers’ perceptions of issues by reshaping the nature of the problems, thereby guiding them to choose corresponding policy-making logics. Problem definition determines the basic ideas and analytical frameworks for policy design, but it is not a purely objective process; rather, it is a process of frame selection. Different problem frames imply differing perspectives on understanding the essence of issues, leading to entirely different solution ideas. Technology companies can enhance the importance, relevance, and weight of specific beliefs on current issues by redefining, constructing, and describing particular policy problems, thereby influencing the content and importance judgments of individual beliefs. In the context of great power competition, technology companies can redefine, construct, and describe issues related to their commercial interests as critical concerns for national security and strategic interests, even persuading decision-makers to examine industrial policies through the logic of national security to seek more policy support for themselves. According to the Copenhagen School’s securitization theory, when an issue is successfully defined as a security issue, decision-makers tend to adopt unconventional emergency measures, breaking existing policy paths and institutional norms. Therefore, technology companies successfully framing commercial issues as security issues will draw decision-makers’ attention to these problems, prompting them to formulate unconventional response policies.
Second, at the policy goal level, technology companies can selectively amplify or downplay the strategic value of different policy goals using the great power competition framework, influencing decision-makers’ judgments regarding priority among goals and establishing those aligned with their commercial interests as higher-priority goals. When formulating policies for specific issues, decision-makers need to weigh the importance, urgency, accessibility, and other factors of multiple goals to determine their priorities and reasonably establish a phased and hierarchical policy goal system. In the context of great power competition, technology companies can amplify the strategic value of policy goals favorable to themselves, emphasizing the critical significance of achieving these goals for national interests and enhancing comprehensive national power to persuade decision-makers to prioritize them. For example, in the field of artificial intelligence governance, technology companies may emphasize the positive effects of a lenient regulatory environment on the country’s ability to seize technological high ground, urging decision-makers to prioritize technological innovation in policy goals. Conversely, for policy goals that may harm their interests, technology companies often adopt the opposite strategy, emphasizing the potential consequences of prioritizing these goals amid international competition, attempting to pressure decision-makers to lower their priority. Additionally, technology companies may exploit the uncertainties of great power competition, exaggerating the complexity and urgency of competitive situations, inflating competitors’ strategic intentions and their pace in key technology fields, creating a sense of urgency to pressure decision-makers to increase policy support in relevant areas.
Lastly, in terms of specific policy content design, technology companies typically seek to downplay unfavorable resource allocation standards or decision-making considerations while introducing favorable allocation standards or decision-making considerations, using the logic of great power competition to justify such behaviors to gain decision-makers’ approval. In the context of great power competition, technology companies attempt to influence decision-makers’ tendencies regarding policy resource allocation by obscuring decision-making focuses and emphasizing geopolitical factors in decision-making considerations to maximize their interests. For instance, domestic technology companies may reinforce the national label in their corporate attributes, making national identity a crucial consideration in decision-making, thereby gaining preferential treatment in resource allocation, receiving lower costs and higher profit margins compared to non-domestic companies. However, from the perspective of policy effectiveness, deliberately downplaying more substantive evaluation dimensions such as innovation capabilities and user feedback may lead to national policy tools deviating from their original intention of incentivizing innovation. Moreover, specific policy content design also involves the selection of policy instruments; decision-makers typically choose and combine various instruments such as taxation, subsidies, and regulations to achieve synergistic effects. For unfavorable policy instruments, technology companies can utilize the logic of great power competition to emphasize the potential negative impacts on national technological competitiveness, prompting decision-makers to abandon or adjust them.
In summary, unlike the technological liberalism framework typically employed by technology companies in the past when lobbying, the great power competition framework provides new frame choices for them. Technology companies with critical positions in technological competition actively adapt to decision-makers’ core concerns, guiding policies to tilt in favor of their commercial interests by reshaping problem nature, altering policy directions, and influencing resource allocations.
3. Case Analysis of Technology Companies Influencing National Policies
The rapid advancement of artificial intelligence technology is altering the existing power dynamics between states and markets, granting major technology companies unprecedented influence. The relative advantages of technology companies in technological development and application are translating into greater bargaining power in their interactions with states. States increasingly recognize that to ensure technological innovation capabilities, enhance government governance efficiency, and secure competitive advantages in comprehensive national power, they must collaborate with artificial intelligence technology companies. In this context, technology companies are actively leveraging the great power competition framework to influence policy-making and achieve their commercial interests. The following will analyze how technology companies navigate these dynamics through representative case studies.
(1) Intel’s Efforts in the U.S. Chip Subsidy Battle
In the context of great power competition, chip manufacturing capability has become a crucial indicator of a country’s technological strength, directly related to supply chain security in key areas. To occupy a technological leadership position in the artificial intelligence era, governments worldwide are committed to providing subsidies to chip manufacturers to develop domestic semiconductor industries, reduce foreign dependencies, and enhance supply chain reliability and risk resilience. To ensure U.S. leadership in technological competition, the Biden administration proposed a chip industry policy centered on government subsidies, with the “CHIPS and Science Act” effective from August 2022 providing a total of $52.7 billion in government subsidies for the U.S. chip industry. Following the enactment of this law, how this substantial subsidy is allocated has become a central concern for chip companies. Consequently, U.S. chip companies have been lobbying in Congress and the Department of Commerce, with Intel being particularly proactive.
As the flagship company of the U.S. semiconductor industry, Intel has long maintained a leading advantage in the central processing unit sector, but its positioning in the artificial intelligence chip field has lagged, especially in advanced process technologies, where the gap between Intel and TSMC and Samsung is widening. Under the leadership of new CEO Pat Gelsinger, Intel is eager to regain its dominance in chip manufacturing, setting a goal to “produce 2nm chips starting in 2024 and introduce five new nanometer nodes within the next four years, surpassing TSMC.” However, the reality is that almost all advanced artificial intelligence chips on the market are manufactured by TSMC. In this context, Intel is attempting to introduce geopolitical factors to secure a favorable position in its policy negotiations with TSMC and Samsung. Gelsinger has repeatedly emphasized that as a domestic U.S. company, Intel should receive more government support. He argues that strengthening the domestic semiconductor industry is not only beneficial for the U.S. economy and jobs but also essential for ensuring supply chain security in critical technology sectors and maintaining the U.S.’s technological leadership. Providing subsidies to external companies could result in funding flowing overseas, undermining the U.S.’s competitive advantage in global chip competition.
In the written testimony at a hearing of the U.S. Senate Committee on Commerce, Science, and Transportation on March 23, 2022, Gelsinger actively influenced the U.S. government’s allocation of funds. First, he emphasized Intel’s uniqueness as a domestic company, stating that Intel is “the only American semiconductor company excelling in smart silicon chips, platforms, software, architecture, design, manufacturing, packaging, and innovative cutting-edge manufacturing capabilities,” and is also “the only one among the three companies globally producing advanced logic chips with most of its R&D and intellectual property based in the U.S.” This highlighted Intel’s importance to the U.S. economy and national security. Secondly, Gelsinger cited authoritative studies indicating that U.S. chip manufacturing is lagging behind East Asian competitors in terms of costs and government support, thus arguing from a national economic security perspective about the dangers of U.S. dependence on East Asian supply chains in critical sectors, emphasizing that “Intel is committed to helping rebalance the global chip supply.” Additionally, Gelsinger added further considerations regarding the allocation of subsidy funds, particularly emphasizing the potential national security impacts of the project and whether it could operate independently of foreign government influence after federal assistance runs out.
Ultimately, Intel’s efforts yielded significant returns. According to the agreement reached between the U.S. Department of Commerce and Intel, Intel received up to $8.5 billion in government subsidies and up to $11 billion in special loan support, far exceeding TSMC and Samsung, making it the largest beneficiary of the “CHIPS and Science Act.” This outcome indicates that Intel successfully guided government policies and resources in a direction favorable to itself by introducing the great power competition framework. TSMC’s founder Morris Chang noted the commercial logic behind this action, pointing out that Intel is leveraging the intense geopolitical competition to amplify perceived national security threats to enhance its competitiveness, and that this approach will create a more complex market environment for TSMC.
(2) Mistral AI’s Lobbying for Regulatory Relaxation in the EU
In the context of great power competition, the competition surrounding artificial intelligence regulation is intensifying. Compared to the U.S. and China, the EU faces numerous challenges in the digital realm, such as an overall relative lag in the digital platform economy, a lack of domestic internet giants, and high dependence on foreign core technologies. To address its hard power shortcomings, the EU aims to establish the world’s first “Artificial Intelligence Act” to dominate global governance rules for artificial intelligence, continuing the “Brussels Effect.” The European Commission, as the primary drafter of the law, advocates for a human-centered approach, promoting risk-based regulation that imposes strict responsibilities and obligations on high-risk applications, including data governance and transparency. However, this user and citizen rights-focused regulatory model often means increased operational costs for technology companies.
As a rising star in the European artificial intelligence sector, Mistral AI has made remarkable progress in developing large language models. Founded in April 2023 by former employees of Meta Platforms and Google DeepMind, the company launched language processing models “Mistral 7B” and “Mixtral 8x7B,” with the latter outperforming similar products in multilingual capabilities and performance. In this context, Mistral AI quickly emerged as Europe’s greatest hope in the global artificial intelligence competition. However, as the EU’s legislative process advanced, Mistral AI grew concerned about the regulatory environment.
To maintain its development space, Mistral AI actively engaged in policy lobbying, attempting to influence legislators’ perceptions and priorities regarding artificial intelligence governance goals. On one hand, the company emphasized Europe’s technological competition disadvantage compared to the U.S. and China, arguing that legislators should reconsider the priority of technological innovation versus regulatory goals. Former French government official and head of Mistral AI’s public affairs, Cédric O, pointed out that Europe’s main issue is not regulation but the lack of leading technology companies that can compete with the U.S. and China, advocating for a focus on nurturing domestic champion companies. On the other hand, the company stressed that regulatory pressures on European enterprises would have negative impacts. Currently, U.S. and Chinese tech giants are heavily investing in artificial intelligence technology, seizing the high ground for industrial development. In contrast, Europe exhibits significant shortcomings in artificial intelligence, and strict regulations could force European companies out of competition, further exacerbating Europe’s lag in the new technological revolution.
Under Mistral AI’s lobbying efforts, subtle changes occurred within the EU. France, a staunch supporter of the law, originally emphasized strengthening governance norms for artificial intelligence systems and vigorously promoted the implementation of trustworthy artificial intelligence systems. However, during the final stages of trilateral negotiations, France, influenced by Germany and Italy, opposed any binding rules for foundational models beyond codes of conduct, seeking mandatory self-regulation for foundational models. Mandatory self-regulation allows regulated entities to establish binding rules, with regulatory agencies approving based on minimum standards and objectives, and having the authority to inspect compliance and take enforcement actions in case of violations. This model may lead to regulatory agencies excessively accommodating the regulated industry, diverging significantly from the EU’s initial push for strict regulation. Under pressure from all sides, France ultimately approved the “Artificial Intelligence Act” but proposed conditions, including ensuring that regulations do not hinder the development of artificial intelligence models, balancing transparency and the protection of commercial secrets, avoiding excessive burdens on enterprises due to high-risk obligations, and reassessing and establishing thresholds and standards for high-risk artificial intelligence models.
It is noteworthy that despite Mistral AI emphasizing its European identity during the lobbying process, attempting to align its development with European interests, the primary driving force behind its actions was not geopolitical factors. Shortly after the law’s passage, the company announced a collaboration with U.S. tech giant Microsoft to launch models on its Azure cloud platform. This move immediately raised questions among European nations regarding its stance. In response, Mistral AI explained that its goal is to develop technologies that meet enterprise needs, thus opting to use Microsoft’s cloud services. Given that generative artificial intelligence applications require substantial computing power for model training and operation, collaborating with corporate giants like Microsoft is a logical choice for Mistral AI.
(3) OpenAI’s Security Discourse in the Technology Closed-Source Debate
As great power competition intensifies, artificial intelligence technology has become a strategic high ground for countries. Large language models, as a key technological pathway in the field of artificial intelligence, exhibit vast application prospects in scenarios such as intelligent dialogue, knowledge Q&A, and content generation. As a pioneer in the development of large language models, OpenAI has sparked a global artificial intelligence frenzy with its flagship products like ChatGPT, establishing its leadership in this field.
However, with OpenAI’s rising influence, controversies surrounding the openness of technology have also intensified. OpenAI originally adhered to an open-source philosophy, promising to share model architectures and training data with global developers to promote the democratization of artificial intelligence technology. However, as the competitive landscape changed, OpenAI gradually adjusted its strategy, emphasizing access control and usage restrictions for its models. In 2019, OpenAI expressed concerns about the risks of model misuse, releasing only a small-scale version of the model parameters when launching the GPT-2 language model. Subsequently, OpenAI announced a $1 billion exclusive investment from Microsoft and granted the exclusive license for its GPT-3 model to the company. The refusal to disclose architectural details in the GPT-4 technical report was a clear signal indicating that OpenAI intended to protect its intellectual property to maintain its advantage in the increasingly fierce global competition.
OpenAI CEO Sam Altman repeatedly emphasized the significant risks associated with artificial intelligence technology in 2023, stating that closed-source practices are motivated by security concerns. He also called for the U.S. to adopt licensing and registration systems for artificial intelligence models exceeding critical capability thresholds in written testimony before the Senate Judiciary Committee. Altman’s remarks regarding security concerns received praise from many U.S. lawmakers but faced widespread skepticism from industry professionals. On one hand, closed-source practices do not guarantee technological security and may even hinder the oversight and improvement of technology. The mainstream view in the machine learning research field advocates open-source culture, believing it facilitates collaboration, accelerates the transition from theory to application, and is crucial for reducing research and development costs and promoting innovation. In contrast, closed-source practices not only limit user choices and competition but also make it difficult for users to scrutinize software’s security, privacy protection, and potential biases. On the other hand, OpenAI has not slowed its pace of technological development due to security risks, nor has it supported strict regulation of artificial intelligence; rather, it has actively lobbied the EU and its member states to loosen regulations during the formulation process of the EU’s “Artificial Intelligence Act.” In fact, there have been internal power struggles within OpenAI, with one faction led by Altman advocating for pursuing commercial profits, while another faction represented by former chief scientist Ilya Sutskever expressed concerns about the immense risks of the technology, hoping to slow down its development. Ultimately, Altman’s faction prevailed during the personnel changes in November 2023, further indicating that the security issues Altman emphasizes are not his core concerns; rather, OpenAI’s closed-source behavior is more of a commercial strategy to ensure its leading position. Considering the high costs associated with developing and operating large models, monetizing technology and creating sellable products are priorities for such companies.
In response to OpenAI’s attempts to consolidate its advantageous position through closed-source practices, some technology companies are competing by releasing open-source large models. Consequently, proprietary model owners are lobbying the government to restrict the release of open-source models by portraying the threat they pose to the U.S. artificial intelligence leadership. In fact, reports in late 2023 indicated that the U.S. Department of Commerce is considering measures to limit the release of future open-source artificial intelligence foundational models, attributing this to alleged “threats from China.” As a U.S.-based technology giant, OpenAI’s strategy of framing technological openness as a security concern aligns closely with the U.S. government’s technology containment policies towards China. OpenAI’s recent decision to allow the military to use its generative artificial intelligence tools further indicates that the company has integrated its technology development with national security strategies. In this context, the U.S. government is likely to adopt stricter regulatory measures based on security considerations to restrict the flow of artificial intelligence technology, thereby consolidating the competitive advantages of domestic enterprises.
4. Potential Risks of Technology Companies’ Proactive Adaptation
As evidenced by the aforementioned cases, amid the competition surrounding technology among nations, tech giants are proactively adapting to secure their advantages in commercial competition. However, such behaviors may threaten the development of technology and industries, ultimately harming national interests and the progress of human society. Specifically, the potential risks can be analyzed from four aspects: resource allocation imbalances, poor regulatory effectiveness, damaged international cooperation, and distorted ecosystems.
(1) The Risk of Resource Allocation Imbalance
The actions of tech giants to influence policy-making driven by commercial logic may lead to imbalances in the allocation of innovation resources. When commercial interests become the decisive factor influencing policy agendas, states may excessively tilt towards applications with short-term commercial value when determining technology development strategies and resource investments, neglecting critical areas like fundamental research that are vital for long-term development.
Although technology companies have propelled the development and application of technology through commercialization in modern society, purely profit-oriented strategies may incentivize companies to prioritize projects that can yield significant economic returns. Currently, there is a trend in the development of artificial intelligence where technology advancements are imitating the technological pathways proposed by OpenAI in 2022; while this accelerates product commercialization, such a model is unlikely to achieve technological leadership or surpassing. Achieving breakthroughs in artificial intelligence requires a profound understanding of existing technologies and accurate predictions of future technological trends, necessitating the allocation of resources and computing power to differentiated technological innovation pathways. In fact, advancements in frontier technologies typically require long-term resource investments and theoretical accumulation. Although many leading technology companies have research teams, their output speed and profit scale often struggle to compete with project teams. Consequently, during periods of economic downturn, research teams are often the first to be reduced or eliminated, or they may choose short-term achievable solutions under performance pressure rather than pursuing more effective but longer-term solutions.
In this regard, NVIDIA’s development strategy provides a model for balancing technological innovation with commercialization needs. Its research team effectively supports the construction of the chip ecosystem and achieves technological breakthroughs, demonstrating the advanced nature of its graphics processing unit technology, making its products popular in the market. One important reason for this is the relatively lenient research and development environment and open corporate culture, where assessments focus more on the value of technology rather than immediate commercial conversion, and where the participation of scientists in most internal meetings is not restricted, facilitating their deep understanding of commercial needs and industry developments, thus aligning their research efforts more closely with market demands and technological frontiers.
In light of this, states need to be cautious about the trend of resource allocation excessively favoring commercial transformation while also considering the long-term technological development needs, guiding innovative entities to allocate resources for short-, medium-, and long-term technological reserves, ensuring sustainable research and innovation capabilities in critical areas.
(2) The Risk of Poor Regulatory Effectiveness
David Collingridge has profoundly elucidated a core dilemma in the field of technology regulation. He points out that society faces a dual dilemma in controlling technology: in the early stages of technological development, while it is easier to control, it is challenging to formulate effective control measures due to a lack of full understanding of its potential social harms. Conversely, when the negative social impacts of technology become apparent, control often becomes costly and slow. Therefore, while pursuing innovative potentials, countries need to recognize the importance of effective technology regulation. A lack of effective regulation may lead to a proliferation of standards and norms, exacerbating issues of technological instability, unreliability, and safety, hindering further technological development and application.
Currently, technology giants occupy advantageous positions in market competition due to their vast data and advanced algorithms. If states fail to prioritize technology regulation under the lobbying of technology companies, it could result in technology companies abusing their dominant positions, ultimately impacting individual privacy, social equity, and market competition negatively. For instance, some companies have exposed biases and discrimination issues when using artificial intelligence algorithms for recruitment, credit, and decision-making. Moreover, many regulatory agencies have noted the role of application programming interfaces (APIs) in the digitization and platformization process. As the common language for data and service exchange between enterprises, APIs have become core elements of digital infrastructure, supporting the development of platform economies and societies. Technology giants like Google and Facebook typically manage different users of their platforms through APIs, leading to ecosystems of applications and services reliant on their APIs that can easily suffer disruptions or chain reactions from minor changes, even becoming paralyzed. More seriously, if the services of technology giants are maliciously manipulated or attacked, the vast amounts of personal and sensitive data they collect and store may leak, threatening social stability and national security.
(3) The Risk of Damaged International Cooperation
In the context of increasingly fierce global technological competition, the lobbying actions of technology giants introducing geopolitical factors may lead to the spread of domestic protectionism, hindering normal international technological exchanges and cooperation. This will undoubtedly exacerbate tensions and confrontations in international relations, increasing market barriers and technological divides, and heightening the difficulty of reaching consensus on technological standards, ethical norms, and regulatory policies among countries, thereby undermining the foundation for international cooperation in addressing common challenges.
It is crucial to be vigilant about the negative impacts of geopolitics on technological exchanges and cooperation. Since 2010, the number of AI research cooperation projects between China and the U.S. has increased by about four times. However, from 2020 to 2021, the total number of cooperation projects in this field grew by only 2.1%, marking the lowest growth rate since 2010. In the face of the excessive securitization of U.S. technology policies, Tobin Smith, senior vice president of the American Council on Education, stated that such practices could undermine the openness of science. Throughout the history of technological advancement, open international cooperation has always been a key factor in promoting technological innovation and the conversion of applications. The development of internet technology itself relies on the collective wisdom of global scientists and the support of cross-border supply chains. Splitting the internet into multiple independent networks at the infrastructure level means replicating a highly complex supply system, which is costly and impractical. Thus, countries need to carefully weigh technological sovereignty against international cooperation, remaining vigilant against intensified commercial competition exacerbating strategic suspicion and confrontation among nations.
(4) The Risk of Distorted Ecosystems
In formulating specific industry support policies, countries need to avoid the risk of technological ecosystem distortion caused by the reinforced monopolistic positions of technology giants. Such distortion may lead to increased research and development costs, delay technological progress, suppress the vitality of entire industries, and ultimately weaken a nation’s comprehensive economic strength and international competitiveness, harming the long-term interests of countries and humanity.
The sustained healthy development of technology relies on a sound ecosystem, which typically comprises multiple interrelated and interdependent entities, including technology research institutions, technology companies, regulatory agencies, investors, and users. In the global technology competition, countries with sound ecosystems can attract domestic and foreign investments and innovators, nurture local technology companies, and enhance their influence in global technological innovation and standard-setting. However, some technology companies, leveraging their monopolistic advantages in funding, technology, and talent, control the upstream and downstream of the industry chain through mergers, patent licenses, and other means. If, with government support, the monopolistic positions of technology giants are further reinforced, the operational space and voice of other actors within the ecosystem will be further squeezed. Once technology giants dominate the formulation and application of industry standards, they tend to maintain the existing pattern and exclude disruptive innovation attempts. This homogenized technological evolution path is unlikely to meet the diverse needs of society and will also weaken the industry’s resilience to external shocks. Over time, issues of insufficient motivation for technological innovation and weakened industry resilience will continue to deteriorate, further impacting a nation’s overall innovation capability and international competitiveness.
Conclusion
The rapid development of artificial intelligence technology is reshaping the global political and economic landscape, and the evolution of the relationship between technology companies and states will determine the future picture of the artificial intelligence era. Currently, tech giants have become the leaders of global technological transformation and key players in great power competition. As the concentration of power among a few large technology companies continues to rise, the technological gap between nations will deepen, posing challenges to global technology governance. In the future, the interactions between technology companies and sovereign states will continue to evolve, and how they balance each other’s goals and interests will become an important entry point for observing global political and economic dynamics.
In the face of the escalating technological competition among nations and the potential risks brought about by technology companies’ involvement in geopolitics, countries need to actively address the power and interest conflicts arising from technological changes. At the same time, they should place high importance on the role of technology companies in this context, guiding them to play a constructive role in technological innovation and enhancing social welfare. The potential of artificial intelligence technology to promote human social progress is immense, but harnessing this disruptive technology requires global cooperation. The vision for the development of artificial intelligence technology should transcend national interests, focusing on the well-being of humanity. This necessitates technology companies to participate in global technology governance in a more open, transparent, and responsible manner, while major countries must strengthen dialogue, jointly explore technological development directions, and cooperate to establish norms that align with global interests, working together to build a more inclusive and sustainable global technological development ecosystem.
Source | WeChat public account “Political and International Relations Analysis,” September 26, 2024.
This information does not represent the views of the platform.
Layout | Gao Hanruo
Review | Liu Zhuowei