Abstract
Generative artificial intelligence represents a significant opportunity, particularly in educational settings, as it has the potential to enhance the teaching, learning process. However, its use should aim to integrate and augment, rather than replace, human thought—adopting a logic of inclusion (et, et) rather than exclusion (aut, aut) (Prencipe & Sideri, 2023, p. 99). Firstly, this approach counters the apocalyptic tendency (Eco, 1964) to reject technology—specifically, in this case, AI—on the mistaken assumption that it substitutes human intelligence and, consequently, the learning process. Instead, AI should be framed as a tool designed to facilitate human learning processes (Prencipe & Sideri, 2023, pp. 100, 102). Nevertheless, a fundamental shift must be acknowledged: learning should transition from a focus on merely identifying solutions to a practice centered on formulating well, structured questions, which, in turn, generate meaningful content—whether in the form of images, videos, text, or other media. This is the space within which educators must operate when teaching the use of generative digital tools. In doing so, they should engage with themes of lateral and divergent thinking (Sibilio, 2014, 2016) and embrace the human creative capacity to envision and construct new possible worlds (Berthoz, 2015; Sibilio, 2017). At the same time, the challenge between the various nations is only just beginning, both in terms of the chips that can be used for computing power, and in terms of performance in the generation of multimedia content through artificial intelligence. That is why it is important for a nation to be among those allowed to use top, of, the, line technologies. In the context of increasing regulations on the dissemination of artificial intelligence (AI), the U.S. Framework for Artificial Intelligence Diffusion, introduced by the United States, imposes restrictions that may impact not only global security and the economy but also access to AI, based educational technologies. By limiting the availability of advanced models and specialized hardware, these measures risk creating a divide between countries with privileged access and those subject to restrictions, potentially affecting education, research, and innovation in the educational sector. The U.S. Framework for Artificial Intelligence Diffusion is a regulatory framework issued by the U.S. Department of Commerce on January 13, 2025, with the objective of managing the global dissemination of advanced AI technology through export controls. This regulatory framework aims to balance the protection of U.S. national security and foreign policy interests with the promotion of economic and social benefits derived from the responsible dissemination of AI. Key measures introduced include global licensing requirements for the export of advanced computing, integrated circuits and the “weights” of the most sophisticated AI models, as well as the implementation of licensing exceptions for low, risk destinations and end users. The regulation officially took effect on January 13, 2025. The U.S. Framework for Artificial Intelligence Diffusion also includes restrictions on the types of advanced chips that can be exported, particularly high, performance processors used for training and executing sophisticated AI models. The United States seeks to limit access to these chips by countries deemed high, risk for national security or potential military applications. Countries classified as “high, tier” (low, risk countries, among which Italy) benefit from licensing exceptions, meaning they can more easily access advanced hardware and the most sophisticated AI models without undergoing lengthy approval processes. This facilitates technological innovation, research, and development without significant restrictions. Conversely, lower, tier countries may face limited access or be required to obtain case, by, case approvals, slowing the adoption of advanced technologies and putting them at a disadvantage compared to nations with fewer restrictions (Clifford Chance, 2025; Rand Corporation, 2025).
The U.S. Framework for Artificial Intelligence Diffusion could also affect AI, related educational software in three main areas: 1) Access to Advanced AI Models (if an educational software application relies on advanced AI models for virtual tutoring, content generation, or personalized learning, it may be subject to restrictions if these models require advanced hardware or neural network weights that fall under export controls); 2) Availability of Hardware for Training (schools, universities, and startups in restricted countries may face challenges in accessing GPUs and advanced chips necessary for research and development in the educational sector. This could limit the adoption of AI, based educational tools or slow down the development of new applications); 3) International Collaborations and Exchanges (if a country is not classified as “high, tier,” obtaining licenses to use state, of, the, art AI software and hardware may be difficult. This could impact global educational projects, such as Massive Open Online Courses (MOOCs) or advanced e, learning platforms that leverage AI). In modern education this will have a strong impact, due to the fact that digital technology plays a crucial role in enhancing learning experiences.
Keywords
NA