The current Generative Artificial Intelligence (GAI) technology is developing vigorously, bringing transformation to network security technology and industry. The application of AI technology has driven the escalation of confrontation between network attack and defense technologies. Network security products and applications based on large model have become the new focus of network security industry, injecting new momentum into the development of network security industry. At the same time, the application of AI large models has also brought new security threats. Countries and regions around the world are continuously paying attention to the network security issues caused by GAI, and have introduced various targeted regulations and measures. Therefore, by analyzing the impact of GAI on China’s network security industry, and proposing countermeasures and suggestions, it is of great significance for enhancing the country’s network security protection capabilities.
The Generative Artificial Intelligence (GAI) model profoundly affects fields such as information dissemination, content creation, and social interaction, but also brings a series of security challenges such as data privacy breaches, false content generation, and intellectual property protection. By exploring the security risks and corresponding governance strategies of GAI at the current stage, this paper hopes to provide reference for the sustainable and healthy development of this technology. Firstly, this paper analyzes various security challenges caused by the development of GAI technology. Secondly, this paper discusses typical security issues of GAI, including data security and privacy, model abuse, algorithm stability, and the reasons why GAI technology can be maliciously attacked in text generation, image recognition, and other areas. Finally, this paper explores the establishment of a multi-level GAI security governance framework, including technical, organizational, and social layers, as well as secure, controllable, and reliable governance strategies and specific governance paths. Through collaborative governance by artificial intelligence enterprises, academia, government regulatory departments, and the public, this paper aims to enhance the overall society’s awareness and response capabilities to GAI security.
Since the appearance of ChatGPT in 2022 and the release of Sora in 2024, generative artificial intelligence has become the focus of public attention once again. As an important part of the new generation of artificial intelligence, generative artificial intelligence is accelerating its integration into the economy and society, promoting the transformation of traditional industrial production methods and the upgrading of circulation paths, creating digital industrial clusters. Generative artificial intelligence gradually shows its catalytic role in the digitalization and intelligent transformation of manufacturing. However, the automatic content generation based on big data training also leads to new cyber security risks, and the problems such as harmful content generation and data leakage are becoming increasingly prominent. Through a comprehensive analysis of the integrated development and security situation, this paper scientifically demonstrates the cyber security risks faced by the application of generative artificial intelligence in manufacturing industry, and puts forward relevant countermeasures.
Based on the generative artificial intelligence, this paper studies the Policy-Protection-Detection-Response(PPDR) model. And this paper effectively relies on the intrusion detection, firewall, honeypot technology, network vulnerability scanning and other advanced technologies of the generated artificial intelligence model, and then builds a scientific and systematic network security dynamic defense model. The model shows strong active defense function, and can show the dynamic and proactive characteristics of network security in an all-round way. The whole protection system of generated artificial intelligence model changes greatly depending on the interaction and cooperation among decoy system, intrusion detection system and firewall. It realizes the transformation from static state to dynamic state, and greatly improves the motivation of firewall, while the overall protection ability of network reaches a high level.
As the application of large language model-driven agents deepens in various fields, potential security risks are gradually prominent. This paper aims to systematically sort out the security and trustworthiness problems faced by agents based on large language models, including information leakage, model attacks, hallucination outputs, ethical and moral risks, and legal compliance hazards. By conducting an in-depth analysis of the causes and impacts of these security risks, this paper discusses existing protective measures and technical means, and proposes suggestions for building trustworthy large language model agents, providing references for related research and practice.
Data security is an important component of national security, which not only concerns a country’s information sovereignty, but also affects social stability and international competitiveness. In the increasingly complex international and domestic environment, the problems faced by data security governance have become more severe. This paper focuses on the analysis of data life cycle risks, and puts forward the general idea of data security system from three aspects: technology, operation, and management, and helps all participants to establish a data security system.
With the continuous upgrading of security needs in the process of industrial digitization, the development of zero trust in China has accelerated. This paper first reviews the zero trust policies and standards, and studies how zero trust solves the security problems faced by enterprises in the six areas, namely, identity, terminal, network environment, applications and workloads, data, and security management. Then, based on the investigation on the supply-side zero trust enterprises and the application enterprises in the financial industry in China, this paper forms the supply capability mapping and analyzes the application scenarios and roles of zero trust. Finally, it analyzes how artificial intelligence empowers the development of zero trust.
Currently, global 6G research is in a highly active period, and all parties are continuously promoting 6G security work. This article firstly reviews the international standardization of 6G security and the progress about key technologies of 6G security. Then, starting from the typical application scenarios of 6G released by the International Telecommunication Union, the security requirements of 6G networks are analyzed. Finally, based on the security requirements of 6G, the key technologies that need to be studied in 6G security are discussed, and the future development and research directions of 6G security are proposed.
As a new network architecture, SDN has been widely recognized in the industry and widely applied and deployed on a large scale. However, SDN has become the main target of DDoS attacks due to its centralized control mode, causing significant harm to SDN networks and related applications. This paper takes DDoS attacks in SDN as the research object. Firstly, the potential risks of DDoS attacks in SDN network architecture are summarized and the main forms of DDoS attacks faced by SDN at present are analyzed. Then, the main solutions for detecting and defending against DDoS attacks in the industry are introduced and the main problems in current research on DDoS attacks are discussed. Finally, the future research work of DDoS attack in SDN is discussed.
With the rapid development of cloud computing technology, more and more enterprises and individuals are choosing to migrate their business and data to cloud platforms. However, the openness and shared nature of cloud environments also bring numerous security challenges, among which access control stands out as a critical component for ensuring cloud security. This paper conducts in-depth research and discussion on the security design and implementation of access control in the cloud environment. It proposes a new system to solve the design and implementation of access control security, which includes the use of blockchain authorization security, user behavior analysis, feature extraction, trust analysis and other auxiliary decision-making to improve the business security of users in the cloud environment.
India puts great emphasis on the development of Artificial Intelligence (AI), and has introduced a series of policies to promote the application and development of AI within the country. In particular, India also focuses on integrating its AI strategy with the “Made in India” initiative, which aims to enhance the level of India’s manufacturing industry. By intensifying the development and application of AI technology, India aims to achieve iterative upgrades in manufacturing standards and a large-scale expansion of the manufacturing sector.
The integration of large AI models with industrialization holds immense potential and has emerged as a pivotal force driving global economic growth. Based on a thorough analysis of the current application situation of large AI models in the industrial sector, this study emphasizes the importance of technology integration and forward-looking planning. It proposes an overall architecture for the industrial application of large AI models, aiming to uncover key technologies and challenges in the integration process, analyze innovative implementation pathways and integration application challenges. Additionally, it presents countermeasures and assesses future trends, aiming to provide guidance for industry practices.
Quantum Key Distribution (QKD) and Quantum Identity Authentication (QIA) are the core of quantum cryptography, providing a new approach to information security. QKD ensures the security of keys during transmission by utilizing the principles of quantum mechanics, while QIA ensures the authenticity and unforgeability of identity authentication through the properties of quantum states. The combination of the two can effectively address the security challenges brought by quantum computing and provide a high level of protection.
With the continuous development of quantum computing and classical computing, hybrid computing which integrates both advantages has gradually emerged. This paper briefly introduces three workflows of hybrid computing and summaries the latest advancements. Then, it discusses three typical hybrid computing architectures that are compatible with current hardware development: batch-based hybrid computing architecture, session-based interactive hybrid computing architecture, and distributed hybrid computing architecture. Finally, it discusses the key factors which influences the development of hybrid computing, and proposes suggestions to further promote the development and application of hybrid quantum-classical computing.