Generative artificial intelligence poses systematic challenges to the traditional human-centrism-based intellectual property system in two dimensions: the fair use of training data and the copyrightability of generated content. It exposes the deep ethical tension between technological innovation and the protection of creators' rights and interests. Adopting a functionalist comparative law approach, this paper systematically analyzes the differentiated regulatory paths regarding the use of AI training data and the copyrightability of generated content in the United States, the United Kingdom, Japan and China. Combined with utilitarian incentive theory and personality right ethics, it explores the ethical limitations of paradigms such as the Absolute Control Doctrine and the Substantial Contribution Doctrine in terms of transparency, fairness and liability attribution. By establishing the theoretical benchmark of "creativity ecosystem integrity", this paper constructs a hierarchical consent mechanism, an ethical review system for copyrightability, as well as an ethical insurance and risk-sharing mechanism. It aims to achieve a dynamic balance between ensuring fair participation of human creators in value distribution and promoting technological innovation, and form a normative framework for global artificial intelligence intellectual property governance.

The inherent tension between trademark rights and freedom of expression presents a persistent challenge in intellectual property law. This paper examines this conflict within the U.S. legal framework, further investigating the pivotal Rogers test established to balance trademark protection under the Lanhm Act with First Amendment speech guarantees. Through a literature review and case analysis, the study analyzes the test's application and evolution, particularly in cases involving parody and social commentary, and contrasts the U.S. approach with those of EU and the UK. The findings shows that while the Rogers test provides crucial protection for expressive works, its application is not absolute and faces significant limits when a trademark is used as a source identifier, as clarified by the Supreme Court's recent decision inJack Daniel's v. VIP Products.The paper concludes that achieving a precise, context-sensitive balance remains essential for the future development of trademark law in a rapidly evolving communicative landscape.

In recent years, velocity-based training(VBT) has garnered growing attention as a novel method due to its ability of track training loads with greater precision and effective training outcomes assessment. The central mechanism of VBT relies on velocity data recorded during movement to assess, adjust and control the training load. 68 male college athletes from the seven sports were chosen as the subjects of this study. After 12 weeks of a strong-training plan, it was hoped that changes would occur in the Load-Velocity relationship of the bench press before and after the program. A German Eonde Pro device was employed to measure the speed of the barbell. Equations for the linear regression of MV, MPV and PV were developed in Origin2024. To compare the above indices for people at different levels of strength, SPSS was employed. Analysis of the speed of different loads for barbells shows that MV, MPV, PV and load are negatively correlated, and no significant differences in movement speed were found among the sports. After 12 weeks of strengthening exercises, the upper-body strength of all 68 people increased substantially; however, the mean values of MV, MPV and PV were not very different. ANOVA was employed to compare the means of MV, MPV, and PV among the seven sports; that is to say, no single sport consistently outperformed the others in all three velocity indicators. The stability of the load-velocity relationship for the bench press in different sports, strength levels and after organised strength training were all verified in the study. The above results can provide a scientific basis for the construction of strength training programs for athletes of various sports.

Existing Creative Commons (CC) licenses present compatibility issues with the framework of China's Copyright Law. This paper employs a literature review approach to explore the practical challenges and optimization pathways for CC licenses in China's open access publishing sector, with a focus on analyzing the difficulties surrounding the interpretation of terms and the allocation of liability. The study finds that CC licenses suffer from issues such as ambiguous interpretations of non-commercial use and imbalanced allocation of liability. These difficulties stem primarily from the structural differences between the case-by-case balancing approach rooted in common law and China's three-step fair use test, as well as the mismatch between the default allocation of liability under CC licences and China's principles of tort liability.Therefore, localized interpretation rules should be established to clarify publishers' legal obligations and improve supporting mechanisms, thereby promoting deeper alignment with China's legal system and achieving a virtuous cycle between open access and copyright protection.

As the world enters the Fourth Industrial Revolution, the demand for highly skilled talent in the context of industrial upgrading is becoming increasingly urgent. However, in mountainous and rural areas, the imbalance between vocational education and economic development remains a prominent issue. This paper focuses on the structural mismatch between vocational education programs in mountainous areas and local industrial upgrading, pointing out that the current contradiction is not merely a shortage of talent, but rather a misalignment between local needs and three key aspects: program categories, the timing of adjustments, and skills development. By examining the characteristics of mountainous regions—such as fragmented and small-scale industries, limited enterprise absorption capacity, scarce educational resources, and labor outflow—the article analyzes why traditional program structures struggle to respond promptly to emerging skill demands in specialized agriculture, ecological industries, cultural-tourism integration, and digital marketing. Based on this analysis, the article proposes establishing mechanisms for regional collaborative planning, forward-looking industry monitoring, interdisciplinary skills development, and more targeted fiscal support to enhance the capacity of vocational education to serve the endogenous development of mountainous regions.

Many influential approaches to AI value alignment rely, at least in part, on empirical inputs drawn from human experience, including individual preferences, collective judgments, and large-scale datasets. These inputs are indispensable for training, evaluating, and coordinating AI systems. However, they are sometimes treated, implicitly or explicitly, as if they could also provide the normative authority of alignment standards themselves. This article challenges that assumption through a meta-ethical analysis. It argues that preference-based, consensus-based, and data-based approaches all encounter the same structural limit: they can describe how people judge, choose, or behave, but they cannot by themselves explain why such judgments, choices, or behaviours should be treated as normatively authoritative. Taking universalizability and justifiability as two minimal conditions for normative grounds, the article examines the strongest versions of the three empirical approaches and shows that each requires an additional normative premise that cannot be generated from empirical inputs alone. The argument does not reject empirical alignment as a practical coordination mechanism. Rather, it repositions empirical alignment as normatively insufficient but epistemically and operationally useful. The normative authority of AI alignment standards must therefore be secured by independent ethical, legal, or institutional justification.

Generative artificial intelligence (AIGC) can now generate text as well as make pictures, music, and all kinds of expressions by itself. So, the problem about how to set up copyright for AIGC produced work becomes an important issue in legal field and actual work field. The traditional copyright law is based on the idea of human creation. But, in fact, it's not created directly by man. It poses major issues on copyright law as to whether these can be considered works of copyright law, who the author is, who has ownership over it, and how one would determine infringement liability. The copyright legal situation of AIGC is studied from the angle of work identity, authorship and owner identification as well as violation liability determination. Based on related parts of China's Copyright Law, it takes reference from the domestic and overseas studies as well as judgments to look for solutions under current laws and gives practical advice with Chinese characteristics. It can provide a theoretical basis to improve the copyright protection methods and promote the healthy development of artificial intelligence industry.