By Tan Tieniu: Department of Research, Studies and International News12–09-2025

On April 25, 2025, General Secretary Xi Jinping pointed out during the 20th collective study session of the Political Bureau of the CPC Central Committee: “Artificial intelligence (AI), as a strategic technology leading a new round of technological revolution and industrial transformation, is profoundly changing human production and lifestyles.” AI technology has greatly liberated human mental capacity, fundamentally transforming the paradigm of knowledge production and reshaping the landscape of human civilization at an unprecedented speed and scale. It is profoundly influencing the global economic, political, and military landscape as well as the development of human society, increasingly drawing high attention from governments, industries, and academia worldwide. In contrast, the adaptation of the AI governance system has significantly lagged behind the breakthrough progress of the technology, making international AI governance issues increasingly prominent. Therefore, we need to accurately grasp the global development trends of AI, scientifically assess the risks and challenges of AI development, strengthen international governance and cooperation, and promote AI development in a benign direction for the benefit of humanity.

Accurately Grasping the Global Development Trend of Artificial Intelligence

Since its inception in 1956, AI has undergone nearly 70 years of ups and downs. In recent years, generative AI technologies represented by ChatGPT, Sora, and DeepSeek have achieved continuous breakthroughs, sparking a new wave of AI enthusiasm globally. Overall, the development of AI features the following basic trends:

1- Specialized AI Is Increasingly Maturing

Specialized AI refers to AI intended for specific fields, with singular tasks, clear demands, and well-defined application boundaries, therefore achieving breakthroughs in respective fields. In single-item tests of partial intelligent levels, it can surpass human intelligence. For example, biometric features such as faces, irises, and gaits can be recognized with high precision and efficiency, and have been successfully applied in multiple fields including law enforcement, civil affairs and disaster relief; AI-assisted diagnostic systems have covered over 3,000 common diseases and are used in more than 27,000 primary healthcare institutions; Tesla’s fully autonomous driving has achieved the first end-to-end AI-driven autonomous driving, among others.

2- Large Models Have Achieved Significant Breakthroughs

Large models refer to machine learning models with ultra-large-scale parameters and complex computational structures. As the most attention-grabbing direction in the current AI field, large models are increasingly becoming strategic infrastructure. Leveraging the development paradigm of “pre-trained large models + fine-tuning for downstream tasks”, training costs and development cycles can be effectively reduced to enable scale production. Since the release of ChatGPT at the end of 2022, large models have entered a period of explosive growth, acquiring capabilities for multi-modal understanding and multi-type content generation. At the beginning of 2025, DeepSeek-R1 emerged, offering advantages such as open-source innovation and low computational costs, making large models truly accessible to the general public.

3- Generative AI Is Thriving

Generative AI refers to AI technology that automatically generates text, images, audio, video, code, and other content based on algorithms, models, and rules. Compared to traditional AI systems that can only process based on input data, generative AI systems can create new content on their own. Driven by large models, generative AI has developed rapidly and is widely applied in professional Q&A, programming, poetry creation, painting, and other areas, demonstrating powerful generative capabilities that have rocked the industry. For example, ChatGPT was included in the Nature’s 10 in 2023, marking the first-ever inclusion of a computer program in the list.

4- Embodied AI and Humanoid Robots Are Gaining Attention

Embodied AI refers to the intelligence seeking continuous growth through the physical entity’s ongoing interaction with the external environment. It can overcome the limitations of traditional AI, which relies on stacking “computational power + data” and lacks adaptability to scenarios, providing a new perspective for exploring the mechanisms of human intelligence formation and a powerful engine for translating scientific and technological achievements into real-world productivity. Humanoid robots, as an important carrier of embodied AI, provide an ideal platform for studying the emergence of intelligence. Today, research on embodied AI has entered a fast track and will further drive the paradigm shift of AI from virtual cognition to physical intelligence through the closed-loop iteration of “data–algorithm–carrier”.

On January 22, 2025, in Davos, Switzerland, UN Secretary-General António Guterres stated at the World Economic Forum Annual Meeting Davos 2025 that climate change and the uncontrolled expansion of AI have been threatening the survival and development of humankind.

5- AI-driven Scientific Research Is Developing Rapidly

Over the past few centuries, scientific research has undergone transformations from experimental observation, theoretical analysis, computational simulation, to data-intensive paradigms. Currently, AI-driven scientific research is leading a new wave of transformation, which will fundamentally change the way scientific discoveries are made. For example, AlphaFold2 can accurately predict the 3D structure of proteins, while AlphaFold3 can predict the structure and interactions of all life molecules with unprecedented precision, bringing revolutionary changes to biomedical research. Additionally, AI has been widely applied in vaccine development, chip design, weather forecasting, and many other fields.

6- AI Is Accelerating Its Empowerment of Various Industries

As a strategic technology leading a new round of technological revolution and industrial transformation, AI is profoundly changing human production and lifestyles. AI is accelerating its deep integration with the real economy, enhancing social productivity, promoting the transformation and upgrading of traditional industries, fostering the vigorous development of emerging industries, cultivating new highlands for future industries, and driving the rapid growth of the “unmanned economy”. Meanwhile, AI is having a significant impact on social life domains such as transportation, home living, education, healthcare, and elderly care, providing better solutions for promoting social equity, enhancing people’s well-being, and improving the quality and happiness of human life.

7- The Social Impact of AI Is Becoming Increasingly Prominent

Along with the rapid development of AI, its social impact is diverse and all-encompassing. On the one hand, AI brings many positive effects, such as boosting the economy, serving people’s livelihoods, and benefiting society; on the other hand, it also brings many social problems, such as security risks, legal uncertainties, ethical anomalies, and privacy breaches. For example, the misuse of generative AI technology to create false information for defamation, extortion, and malicious attacks poses significant security risks; the ability of smart devices and algorithms to explore and analyze personal privacy far exceeds that of previous technologies, making personal information and privacy issues including personal trails, consumption data, biometric features increasingly complex and difficult to prevent.

8- General AI Remains a Long-term Goal

General AI aims to develop machines with the same level of intelligence as or even higher than that of humans, and may even enable machines to have autonomous consciousness. In recent years, as the capabilities of neural network models continue to improve, the public has developed overly high expectations upon AI, and some experts and scholars have made unrealistic predictions about the development of general AI. Currently, AI still has a significant gap from the level of human intelligence, with obvious deficiencies in semantic understanding, reasoning, planning, generalization, and continuous learning. It also lacks an essential understanding of the spatiotemporal operational laws of the objective physical world and lacks deep empathy. There is still a long way to go before achieving true general AI.

 On March 25, 2025, at the Boao Forum for Asia Annual Conference 2025 held in Hainan, a group of robots showcasing cutting-edge technology became the focus of attention.

Scientifically Assessing the Risks and Challenges of AI Development

General Secretary Xi Jinping pointed out: “AI brings unprecedented development opportunities and also unprecedented risks and challenges.” While AI accelerates the profound integration of the digital and physical worlds, it also blurs the boundaries between the virtual and the real to some extent. Its inherent uncertainties, opaqueness, and lack of explainability are becoming increasingly prominent, posing numerous risks and challenges to human society.

1- Risks and Challenges to National Security

The first is ideological security risks. Erroneous ideological trends such as historical nihilism and populism may spread in cyberspace through algorithmic recommendation technologies, impacting ideological security. The second is national governance environments risks. AI can create false information and confuse public perception through deepfake technologies, and use precise algorithms to exacerbate the “information cocoon” effect and enable opinion manipulation, leading to a weakening of national explanatory power. The third is military security risks. Once a technological “generation gap” is formed in military intelligence, technological powerhouses are more likely to engage in political blackmail and technological surprise attacks against developing countries. The cross-domain military applications of AI in nuclear and space domains will also pose significant risks to military security.

2- Risks and Challenges to Industrial Development

The maturity of AI’s technical capabilities and product forms, the determination of core application scenarios and the degree of industrial integration are all closely related to the development of industries. Achieving mature commercial applications is not something that can be accomplished through short-term massive investment in funds and resources; the path is still long. Taking ChatGPT as an example, its initial investment cost was about USD 800 million, with daily electricity costs of approximately USD 50,000 and per-answer costs of a few cents. Faced with such high costs, for one thing, resource monopolization issues are becoming increasingly prominent, with tech giants forming “digital hegemony” that continuously squeezes the survival space of small and medium-sized enterprises. For another, many enterprises and startup teams are rushing in, and some governments are also eager to follow suit. Blindly rushing in has rendered the development of AI disordered and resource-wasting. Additionally, some enterprises exploit the AI trend for speculative trading, exacerbating market bubble risks. 

3-  Risks and Challenges to Social Governance

The first is data privacy leakage risks. The technical characteristics and application scenarios of AI determine that it can easily access large amounts of private information when being used, potentially collecting personal trails, browsing records, consumption data, biometric features, etc., without users’ knowledge, posing severe challenges to personal information and privacy protection. The second is malicious attack risks. Criminals use deepfake technologies to create false videos and images for extortion, malicious attacks, and other activities, threatening social security and stability, personal reputation, and property. The third is information pollution risks. The misuse of AI technologies has led to the proliferation of false and low-quality content on the internet. This “pollution” of information distorts models’ perception of reality, causing irreversible defects in results.

4- Risks and Challenges to Ethics and the Law

First, the development of AI technologies may trigger significant changes to social employment structures. AI can autonomously complete highly complex tasks, greatly reducing the barriers and costs of content production and interaction. All industries face the risk of labor displacement by AI, which may cause unemployment panic and social disorder. Industry boundaries will further blur, disrupting the long-standing social division of labor system. Second, algorithmic bias and machine discrimination may exacerbate. AI algorithms inherently have limitations in transparency and robustness, which will exacerbate biases or discrimination in generated content, amplifying social injustice. Third, technological ethics issues may be triggered. Technological ethics issues such as digital twins, brain-computer interfaces, and human-machine symbiosis involve risks to human life safety, personal privacy, accountability, social fairness, and justice. The widespread application of humanoid robots will also bring many new issues to a social system where humans are the sole subjects, such as the legal status of “AI companions”. Fourth, the legal system face the risk of failure. The legal subject status of AI systems is questionable, and the attribution of rights and responsibilities is unclear, leading to problems that traditional laws can hardly regulate. For example, the intellectual property rights of AI-generated content are difficult to define, traffic laws for autonomous driving systems still need improvement, and the allocation of liability for losses caused by AI is hard to determine, among others.

Accelerating International Governance and Cooperation on AI

In recent years, the subjects and structures of global AI governance have become increasingly diversified, exhibiting an overall trend of comprehensive domains, multi-levels, and broad cooperation. Various initiatives, legislation, and rules have emerged in a surge, covering politics, economy, law, healthcare, social life, and other fields. Under the framework of the United Nations, countries and various international organizations have actively implemented AI governance and regulation, while industry organizations and enterprises have also participated in promoting self-regulation. Regional cooperation, multilateral cooperation, and bilateral cooperation have continued to strengthen. In 2023, China issued the Global AI Governance Initiative, proposing constructive solutions to widely concerned issues of AI development and governance, providing a blueprint for relevant international discussions and rule-making, and demonstrating China’s sense of responsibility as a major country.

However, current international AI governance still faces multiple challenges. The first is the lack of a unified global governance framework. There is still a lack of consensus on major issues among countries, and many regulatory differences are difficult to coordinate. Non-binding international initiatives have limited effectiveness, and the construction of governance rules with substantive binding force still needs time. The second is divergent interests among countries. Developed economies are more concerned with technological security and ethical issues, while emerging and other economies emphasize economic and social development opportunities. The imbalance of governance power further exacerbates the divergence of focus. The third is the fragmented governance. The rapid development of AI technology and the lag in the construction of the governance system create a contradiction. Static regulations can hardly address dynamic risks, which means a fragmented governance framework with low efficiency.

President Xi Jinping emphasized the need to prioritize people and ensure AI serves good causes, strengthening AI rules governance within the framework of the United Nations. Specifically, accelerating international governance and cooperation on AI requires breakthroughs in the following four areas:

1- Upholding “People-Oriented” Principle and Adhering to the Correct Direction for AI Development

No matter how technology evolves, we must stay true to our original aspirations and pursue benign development in the intelligent era. On the one hand, multilateral governance entities should further reach a value consensus on AI development, establish fundamental governance principles, and aim to enhance human common well-being and serve people’s comprehensive development on the premise of ensuring social security and respecting human rights and interests, ensuring that AI always develops in a direction beneficial to human civilization. On the other hand, the international community should form a risk consensus on AI development and provide guidance at the legal and ethical levels, establishing an AI risk assessment and early warning system as soon as possible to ensure that AI research and development and its application activities are safe, controllable, and aligned with shared human values, preventing harm to national security, public interests, and the legitimate rights and interests of organizations and individuals.

2- Adhering to “Openness and Sharing” Principle and Promoting Inclusive and Accessible AI Applications

Due to differences in scientific and educational capabilities, infrastructure, industrial structures, and socio-economic foundations among countries and regions, the issue of imbalanced AI development has come to the fore. On the one hand, we should deepen open-source initiatives, promote the sharing of AI technologies, talent, and infrastructure, help Global South countries strengthen their technological capabilities, establish an international AI technology sharing platform, break data barriers and technological monopolies, and enable fair access to data resources and shared AI benefits, so as to bridge the global AI divide. On the other hand, we should continuously expand exchanges and cooperation among relevant academic organizations, enterprises, and industry associations, building an open and win-win international scientific and technological cooperation ecosystem, striving to break restrictions imposed by certain countries on scholars from other countries to participate in international exchanges and cooperation, and avoiding technological blockades and talent protectionism that hinder technological exchanges and progress.

3-  Adhering to “Governance Through Technology” Principle and Innovating AI Governance Methods

The decision-making processes and logic of AI are opaque, exhibiting “black box” characteristics. Current technologies lack mechanisms for self-regulation and control, creating a clear governance vacuum at the technical level. Research and development entities in various countries and regions should continuously improve the explainability and predictability of AI, creating auditable, supervisable, traceable, and reliable AI technologies. They should treat computational power, the most intuitive and quantifiable metric, as the object of governance, indirectly achieving AI governance. At the same time, they should use AI technologies to prevent AI risks, actively designing and developing technologies and applications that can effectively govern AI, tracking and evaluating the implementation of AI governance frameworks, and enhancing the technical capabilities of AI governance.

4- Adhering to “System Integration” and Building a Multilateral AI Governance Framework

Against the backdrop of diverse challenges to world peace and development, the international community should build broad consensus globally, actively construct an open, fair, and effective multilateral governance mechanism, and jointly promote AI technologies to truly benefit humanity. On the one hand, countries and regions, as well as relevant international organizations, should strengthen the alignment and coordination of governance rules and technical standards, reach consensus on key issues such as AI safety, ethics, data privacy, cross-border regulation, and military applications, establish and improve laws and regulations, and refine AI ethical guidelines and accountability mechanisms. They should fully respect national differences, enhance the voice of developing countries, and ensure equal rights and opportunities for all countries to develop and utilize AI technologies. On the other hand, other diverse governance entities such as enterprises, research institutions, social organizations, and individuals should actively play roles matching their own identities, jointly participating in the construction and improvement of the AI governance system, achieving the development of AI across the globe featuring extensive consultation, joint contribution, and shared benefits.

International governance of AI is an issue of the times concerning the promotion of building a community with a shared future for mankind. Facing the opportunities and challenges brought by AI, we need both mutual standard recognition and joint risk prevention, as well as ethical guidance and legal safeguards. We must further build consensus and deepen collaboration, always adhering to the “people-oriented” value orientation, ensuring that technological progress keeps pace with civilization development, and making AI a positive force for enhancing human well-being.

——————————————

Tan Tieniu is Academician of the Chinese Academy of Sciences and Secretary of the CPC Nanjing University Committee