Engineering a Sustainable Future Through the Integration of Generative AI in Engineering Education

Abstract

Engineers’ work impacts society and the environment and plays a central role in delivering on the United Nations Sustainable Development Goals. However, developing sustainability skills in engineering programs competes with a dense technical curriculum and has proven challenging. The mainstream adoption of generative AI (GAI) tools has prompted a review of teaching and learning, with expanding possibilities as new use cases emerge. This study reviews the impact that GAI is having on engineering education and proposes a framework for the use of GAI to facilitate greater socio-enviro-technical integration in the engineering curriculum. Based on a scoping review of the literature and a conceptual analysis, this paper provides a forward-looking perspective. Artificial intelligence (AI) is also transforming the practice of engineering, triggering the need to adjust graduate attributes accordingly. The increased productivity expected with the rise of AI in the workplace can scale-up the impact of engineering developments and underscores the need for graduates’ sustainability skills. Furthermore, engineers have a prominent role in the development of AI systems. Therefore, in advocating for the need to enhance graduate’s sustainability skills, we emphasize understanding its limitations and the sustainability of AI systems to address the paradox of AI for sustainability and the sustainability of AI itself.

1. Introduction

Engineering activities shape the world we live in, playing an important role in addressing the United Nations Sustainable Development Goals (SDGs) as outlined in the 2030 Agenda The proposed objectives have a special focus on human-centered sustainable growth, which emphasizes the relevance of the social dimensions of the engineering profession. Consequently, several initiatives have emerged to review the engineering curricula for stronger alignment with the evolving socio-environmental needs, including sustainability and a circular economy.

For example, the Australian Circular Economy Ministerial Advisory Group [1] (DCCEEW 2024) recently released a report with recommendations to transition to a more circular economy, which highlighted the importance of preparing the emerging workforce and the central role that engineers play to build the necessary capacity, including sustainable design, materials innovation, and life-cycle analysis [2]. This reflects a broader ongoing trend of embedding ethical and socio-environmental considerations into engineering education, a concept that has been discussed for decades [3,4] but remains challenging.

Engineering curricula are characterized by a robust and comprehensive technical foundation. As a result, students may perceive engineering as a purely technical field, undervaluing the importance of skills related to ethics, sustainability, and humanistic perspectives [5]. In the case that educators recognize the significance of these dimensions, they struggle to incorporate them within an already dense curriculum. Efforts to interweave socio-environmental elements across various subjects have had limited success, primarily due to resistance against reducing technical content to accommodate these aspects [6].

Sustainability, in particular, has proven difficult to embed meaningfully into engineering programs. Attempts range from introducing standalone sustainability courses, which tend to be general and disconnected from practical engineering applications, to integrating sustainability concepts into existing subjects, which is still work in progress for the effective development of sustainability skills [7], inclusive of the necessary foundation in ethics, and socio-environmental values, aspects often overlooked in engineering programs. The challenges in integrating sustainability concepts highlight a broader issue: the difficulty of achieving a balanced curriculum that meets diverse stakeholder demands, including those of individuals, society, and the environment.

At the same time, the advent of generative AI, particularly tools like ChatGPT, offers new possibilities for transforming teaching and learning. Generative Pre-trained Transformer (GPT) applications leverage advanced natural language processing to provide content ranging from responses to simple queries to complex dialogues. While their rise has sparked concerns about academic integrity [8], they also hold significant potential to induce important changes in engineering education. This article focuses on Generative AI (GAI) due to its ability to go beyond merely providing recommendations or results and for its creative potential, diverse applications, and prominence across all levels of education.

There is a growing number of studies proposing a range of uses of GAI in engineering education and raising awareness on GAI’s anticipated transformative impact. However, an assessment of the scope of this impact and an analysis to broaden engineering education beyond the current traditional curricula for the betterment of the profession remain largely unexplored. The present study addresses two questions: (1) How is GAI impacting engineering education? and (2) How can GAI be used to facilitate greater socio-enviro-technical integration in the engineering curriculum?

In line with the work of authors such as Diaz [9], who views engineering education as a continuously evolving field, the present paper invites a rethinking of educational priorities, a reassessment of the competencies required for future engineers [10], and a consideration of how these competencies should be assessed. As discussed in a preceding work [11], GAI tools’ ability to automate certain tasks increases productivity. While this may allow speeding up the learning process and the engineering throughput, going down that path poses sustainability challenges. Our analysis supports the argument that the productivity gains can be used instead to free up curriculum space, enabling educators to focus more on higher-order cognitive skills and the human, environmental, and social aspects of engineering. This is particularly important given that AI’s capacity to augment engineering projects emphasizes the need to graduate engineers able to assess the impact of those projects in society and the environment. The paper provides a forward-looking perspective to guide educators, researchers, and GAI use case developers, proposing a plausible scenario rather than presenting a comprehensive description of the state of the art.

We underscore the role of engineers in designing responsible applications of GAI and the need to equip graduates with the necessary skills. In arguing for the need to enhance graduates’ sustainability skills, we emphasize the importance of including the ability to assess the sustainability of GAI systems as well, to address the paradox of AI for sustainability and the sustainability of AI itself (Section 8).

The rest of the paper is organized as follows. Section 2 describes the methodological approach, which includes a review of the impact of previous innovations in engineering education (Section 3) and a scoping review of the literature (Section 4). Section 5 discusses approaches to address engineering students’ assessments and suggests areas for further research. Section 6 summarizes the findings and organizes the results in a taxonomy of the impact of GAI in engineering education. Section 7 proposes a framework for the use of GAI to facilitate socio-enviro-technical integration, followed by conclusions, including limitations.

2. Methodology

We conduct a brief historical review of the impact of technological innovations in engineering curriculum and the ways in which new technologies have been incorporated in the engineering curriculum previously. The findings identify a range of approaches that provide background to embed GAI in engineering education first and, second, to reflect on its impact on engineers’ graduate competencies. With these approaches as framework, we conduct a scoping review of reported ways in which GAI is being used in engineering education (

Table 1. Search parameters for this scoping review.

Database	Limits	Search Terms
GoogleScholar	Research articles Years (2023–2025) English language	“AI tools” OR “Generative AI” OR “Generative Artificial Intelligence” OR “AI-assisted” OR “ChatGPT” AND (“Engineering Education” OR “Engineering Faculty”)

). It is worth noting that a preliminary search revealed that when “sustainability”, “circular”, or “sustainable” were included as search terms, no results were found, indicating a gap in the literature. In order to answer the question of how GAI is impacting engineering education, we arrange the findings in a taxonomy (Section 5).

Given the evolving state of this field and the range of possible scenarios, the goal of this study is to influence future developments rather than providing a detailed description of the current state. This explains the choice of a scoping approach rather than a systematic review of the existing literature. The review is not exhaustive, yet it provides an overview that outlines current use cases, trends, and potential, which triggers reflection on responsible use for enhanced sustainability in engineering education and practice.

In order to address the second question (how can GAI be used to facilitate greater socio-enviro-technical integration in the engineering curriculum?), we expand our previous work on the use of artificial intelligence to humanize engineering education [11] following a conceptual research approach to depict a plausible future [12]. The analysis brings together emerging concepts on the need to adapt engineering programs for graduate employability in workplaces that embrace the use of AI, the imperative to enhance sustainability skills in the engineering curriculum in the context of increased productivity, the ability of GAI to realize efficiencies in teaching and learning, and the opportunity to use these efficiencies to facilitate greater socio-enviro-technical balance.

3. Revisiting Core Competencies for Engineers in the GAI Era—A Brief Historical Review of Adoption of New Technologies in the Curriculum

The ways in which previous technological innovations have been integrated in the engineering curriculum provides background for the discussion about the integration of GAI in engineering education and the way in which its integration affects graduates’ expected competencies. The evolution of engineering education is often traced in parallel to the industrial revolutions as shown in Figure 1, ranging from the First Industrial Revolution based on mechanization to the Fourth Industrial Revolution (cyber-physical integration, AI, advanced materials), and more recently Industry 5.0 and Society 5.0, which emphasize sustainability and AI-powered systems. This progression is characterized by the growing integration of technology and sustainability in industrial practices. In parallel, the evolution of engineering education saw a technical and applied focus, gradually incorporating more scientific content to align with the increasing complexity of industrial processes. Over time, accreditation processes became essential for assessing professional outcomes. By the late 20th century, the focus on engineering design shifted to a more holistic approach integrating sustainability, social, and environmental considerations, which has triggered professional accreditation bodies’ review of the expected competency standards [10].

This review of industrial revolutions and their influence on engineering education reveals a recurring pattern: advancements in science and technology consistently drive shifts in how engineers are trained [13]. These transformations are nowadays occurring with greater frequency, following an accelerated pace of innovation that may be approaching the so-called technological singularity, where the technological growth becomes uncontrollable and irreversible, resulting in unforeseeable changes to mankind [14].

With GAI as a very relevant technological innovation that becomes an integral part of engineering practice, educators will have to rethink the core competencies and how they are taught in engineering programs. While some recent publications are still discussing the shifting from traditional engineering education toward a competency-based approach [15], the widespread adoption of GAI forces rethinking the core and professional competencies that should be developed and assessed.

From the point of view of the adoption of new technologies, past technological advancements, such as the following cases, illustrate common approaches.

• Technology replaces manual practice—Technical drawing: A traditional competence of engineers based on manual drawing with ink and rulers has evolved into a modern approach based on CAD tools that are now the standard, with minimal emphasis on freehand drawing for sketches. This change found resistance in its origin based, for example, in the alleged educational value developed by manual drawing.
• Technology as support after learning—Equation solving: Traditional manual methods are still taught as part of mathematics programs, often without online tools. However, nowadays, numerous software tools can solve equations instantly, yet traditional solving is still emphasized at the introductory level because it is considered essential to comprehend the basic scientific principles.
• Technology for augmented skills—Structural or installation calculations: Generally, these are taught following a traditional approach focused on manual calculations for basic structures but resorting to simulation software for higher levels of complexity and real-world scenarios.
• Technology to support learning—Augmented reality learning environments are used to help students understand complex concepts and principles but not allowed during assessment [16].

These examples are chosen to practically represent different approaches to integrating, or excluding, computational aids that precede GAI in the educational process. The first has completely adopted the new ways of drawing virtually replacing manual graphical representations, with new technology replacing old ways. The second mostly excludes the use of external aids as calculators or online calculation tools until competency is demonstrated and allows them thereafter, and new technology is used once mastery is demonstrated. The third opts for a hybrid version where the fundamentals are taught to a certain level. Once these principles are considered as soundly acquired, the use of technical aids is allowed, augmenting the student’s ability. In this case, the competent use of the aid is expected and considered part of the subject. There is also a fourth approach, where new technology is utilized to support learning. Once mastery is achieved, the student is expected to demonstrate their knowledge and skills independently. In this case, the technology is used as scaffolding to facilitate learning.

The first approach would fully embrace GAI, replacing traditional systems with the new technologies, avoiding to some extent the contents and tasks that can be completed by the GAI tools. By doing so, perhaps, it aligns education with professional practices where these tools will dominate, in a way emphasizing efficiency and real-world applicability. However, this method may risk bypassing the foundational skills that are essential for understanding the underlying principles of the tools being used and the problems being solved. Students might become proficient users of GAI but lack the deeper conceptual understanding necessary for innovation or troubleshooting when those tools fail or are not available.

The second approach takes the opposite stance, curtailing the use of external aids, such as GAI, within the educational process. This method prioritizes a purist approach to learning, focusing on developing fundamental skills without reliance on technology. This strategy may cultivate a strong foundation in problem solving and critical thinking, but it risks rendering students ill prepared for real-world scenarios where such tools are or will be ubiquitous. This disconnection from professional environments may hinder students’ ability to apply their knowledge effectively in practice. At the same time, the students’ learning outcomes would be limited by missing out on the benefits of the simplifications, efficiencies, and augmentation that GAI offers. Adopting this position implies imposing restrictions to make it challenging for students to utilize GAI tools and relying on software designed to detect and prevent their use [8,17]. However, such measures risk being short-lived and ineffective as the rapid advancements in GAI technology are likely to outpace detection methods and enforcement strategies. In fact, historically, attempts to ban emerging technologies in higher education have been destined to fail [18].

The third approach represents a middle ground, adopting a hybrid model. In this method, students first master the fundamental principles without external aids to ensure a solid conceptual understanding. Once these skills are established, technology and software are integrated into the curriculum as tools to enhance learning and problem solving. This approach acknowledges the importance of both foundational knowledge and the practical use of modern tools. However, successful implementation of this model requires careful curriculum design to balance these elements effectively, which comes with challenges. First, it is essential and difficult to determine the extent to which foundational theory should be preserved to maintain a deep, conceptual understanding before incorporating GAI into the learning process. Foundational knowledge forms the basis of critical thinking and problem-solving skills, ensuring that students can comprehend and apply principles beyond the scope of automated tools. Second, once the knowledge is established, the reduction in cognitive load achieved using GAI tools frees up mental space that can be used to target students’ development of higher-order skills, such as critical problem solving, creative innovation, and responsible decision making. The decision on what higher skills and competencies to develop is another key consideration. In the context of this article, the interest is focused on socio-environmental awareness, including sustainability and circular design.

The fourth approach encourages the use of the technology during the learning process and expects that students will be autonomous and not use it once the learning has occurred. If GAI is used with this approach, it becomes a learning facilitator but educators must ensure that students do not become over-reliant on GAI while still gaining proficiency in its use. This approach aims at delivering learning efficiencies that may also free up space in the curriculum.

With these approaches in mind, we undertake this literature review and summarize the findings in the following section.

4. The Many Uses of GAI in Higher Education Engineering Programs

Predicting the future applications of GAI within an engineering program is challenging, given the vast diversity of tasks involved and the rapid evolution of GAI systems themselves. Nevertheless, certain applications have already been identified and researched in the literature.

The findings of this scoping review, synthesized in this section, provide insights into current and predicted uses of GAI in higher education engineering programs, including students and faculty views on adoption. In terms of the use that educators can make of GAI, the tasks described in the literature include research, teaching and learning, and curriculum design [19]. While some applications of GAI, like drafting a paper, are relatively straightforward, others, such as utilizing synthetic data or designing survey tools, require only slightly more effort to accomplish. Similarly, in the realm of education, tasks like crafting problem sets, outlining instructions for group activities, and developing syllabi are all feasible applications of GAI [20], even designing course learning outcomes [21]. Many of these tasks can be achieved with basic prompts. As best practices for effective prompting continue to evolve, the outputs are expected to become increasingly precise and tailored.

In relation to the impact on students, most of the surveyed research explores how GAI, especially tools like ChatGPT, can enhance the learning experience [22], including interactive and immersive experiences, enabling students to apply theoretical knowledge in real-world scenarios [23]. Another focus of study is the improvement of students performance by providing GAI assistance in writing code, content comprehension, and generating content outlines [24], followed by applications to enhance learning outcomes and increase students’ engagement [25].

Ghosh (2023) [26] abords the topic of equity and inclusivity, pointing out the importance of ChatGPT as a tool for creating equitable access, especially for international students or students from under-represented backgrounds. His study highlights how GAI can counteract inequitable practices in engineering education that may affect students based on race, gender, ability, or academic background and by creating more inclusive learning environments.

However, students tend to use GAI in their own way. In order to find out the real uses of GAI in the engineering studies, Vidalis et Subramanian [27] performed a survey where student answers included the following: GAI provides quick information tools, helps explain/understand topics better, reduces time in studying and problem solving, provides references for new idees, creates better draft papers, checks work, and facilitates ideas for research. As GAI tools continue developing, the use cases expand, with recent examples of use including idea generation, brainstorming, and overcoming non-technical expertise gaps [28]. Existing research has examined not only how students use GAI but also their perspectives and ethical concerns [29].

To a lesser extent, some studies have also explored faculty perspectives. A qualitative analysis of semi-structured interviews conducted at a U.S. university [30] revealed that faculty members were generally hesitant to integrate GAI into their courses and curricula. Those who did incorporate GAI adopted varied approaches, ranging from allowing students to use GAI individually to actively teaching them how to use these tools ethically and effectively. The study also concluded that faculty role identities, divided in educators, career developers, mentors, applied learning specialists, and facilitators of autonomy, shaped their willingness to integrate GAI into courses. Educators were more hesitant, while career developers encouraged its use. Faculty members hesitant to adopt GAI cited concerns about their lack of knowledge and understanding of the ethical implications and potential impacts on student learning. Many highlighted the need for additional training on the educational use of GAI to make more informed decisions about its implementation. Despite these reservations, most faculty members acknowledged GAI’s growing role and the importance of finding ways to introduce this technology into engineering education.

Guillen and Hernandez (2024) [31] explored the perspectives of both students and professors on using GAI tools in engineering programs at a private university in northern Mexico. The study found that students were more experienced users than their professors, highlighting the need for teachers to become more familiar with these technologies. Professors expressed ethic concerns, but basically about plagiarism, the use of AI-generated content, and misinformation, often holding a negative view that students also perceived. The study’s recommendations focused on addressing ethical issues, particularly around intentional and unintentional plagiarism, and the need for clear guidelines. The proposals encouraged professors to use and evaluate popular GAI tools to better mentor students and incorporate these tools into academic activities with proper oversight. At the same time, significantly, in line with a previous point in this article, additional recommendations included implementing clear rules for GenAI use, improving plagiarism detection, and enforcing sanctions for misuse.

This literature review shows also that the uses of the GAI in education pose the concern of compromised ability to assess the level of student’s attainment, which is difficult to ascertain when GAI tools are used. This situation has triggered discussion and research on rethinking evaluation methods.

5. Rethinking Evaluation in the Age of AI

Traditional methods of evaluating engineering students often exclude the use of internet-connected tools or computers. However, the way in which GAI tools are being intertwined in the engineering curricula makes it difficult to untangle them during students’ assessment. This triggers debate about whether GAI should become a standard component of the engineering toolkit used during evaluations as well. In contrast, another perspective advocates for assessments that deliberately exclude GAI, focusing instead on fundamental problem-solving skills and the foundational knowledge essential to the engineering profession [32].

These contrasting viewpoints highlight the need to balance technological proficiency with a strong theoretical foundation. The proponents of incorporating GAI into evaluations argue that this approach aligns more closely with the realities of professional engineering practice, where using all available tools is the norm. According to this view, evaluations that include GAI can better prepare students for real-world scenarios, where effective problem solving often involves selecting and using the right technologies. Additionally, allowing GAI in assessments could foster a deeper understanding of its applications and limitations, teaching students how to differentiate between effective and poor implementations of such tools. This would not only enhance their technical skills but also develop their critical judgment.

However, integrating GAI into evaluations presents several challenges. One major concern is the risk of students becoming overly dependent on AI, potentially undermining their grasp of the fundamental principles underlying engineering concepts. This reliance could limit their ability to solve problems in scenarios where GAI tools are inaccessible or inadequate. Another critical issue is the difficulty in designing assessments that accurately evaluate critical thinking and practical knowledge, even when GAI assistance is allowed. In many cases, GAI tools like ChatGPT can solve traditional assignments with minimal critical thinking from students. This poses a challenge for educators to create adequate assessments. Further research is needed on how to engage students with their learning, demonstrating their ability to analyze, interpret, and apply knowledge independently, rather than simply relying on AI-generated solutions. Designing this kind of assignments is not trivial and the integration of GAI into engineering evaluations requires a nuanced approach that considers both the opportunities and risks.

To design effective evaluations in the age of GAI, educators must rethink traditional approaches, incorporating strategies that not only reflect the capabilities of GAI but enhance critical thinking, collaboration, and real-world problem-solving skills while safeguarding the integrity of foundational learning. This balance is essential to produce engineers who are both technically proficient and capable of independent, critical thought.

Empirical research is needed on strategies for effective evaluation that test approaches that include the following:

• Focus on application and analysis: Rather than relying on memorization or straightforward problem solving, assessments should center on projects that demand AI-supported research, analysis, and interpretation. For example, assignments could require students to use GAI tools to analyze large datasets, draw meaningful conclusions, and evaluate the implications of their findings. This approach reinforces the importance of not just using GAI but understanding its outputs critically.
• Case studies and simulations: Real-world scenarios are invaluable for teaching students to navigate complexity. By presenting case studies or simulated engineering challenges, educators can require students to employ GAI in solving intricate problems. These exercises would also involve justifying their methodologies and decisions, fostering an appreciation for the nuanced application of GAI in practical contexts.
• Collaborative projects: Team-based assignments encourage students to use GAI as a complement to their collective expertise. These projects can highlight AI’s role as a tool, with students dividing tasks to leverage both human skills and technological capabilities. This approach mirrors the collaborative environments of professional engineering, where interdisciplinary teams work together to solve problems.

These considerations partially imply a redefinition of the educator’s role, such as that noted by Stankovski et al. (2024) [33], a transformation that is neither straightforward nor easy to achieve. Implementing such strategies requires a shift in how educators approach both teaching methodologies and the design of assessments, necessitating adaptability, innovation, and a willingness to embrace new paradigms. A deep understanding of GAI capabilities is essential, as is creativity, in crafting assignments that integrate theoretical knowledge with practical application [30]. Educators must anticipate how students might use GAI tools and ensure that assignments remain challenging and relevant in this new context.

To achieve this, continuous professional development is important. Most of the proposals presented in the literature are student-centered, but up-skilling programs for teachers at technical and pedagogical levels are scarce. Educators need training in elaborating precise and effective inputs for large language models to keep up with students on the use of GAI tools.

6. The Effect of GAI’s Integration in the Engineering Curricula

After reviewing the primary proposed uses of generative GAI, it is apparent that GAI’s impact on engineering education is already noticeable and signals further consequences in multiple fronts. To summarize the anticipated impact of GAI on engineering education, we have created a taxonomy that builds on the findings of the scoping review and classifies the impact in three clusters, curriculum content, cognitive processes, and pedagogical methods, including assessment and structure (

Table 2. Taxonomy of impact of GAI on engineering education.

Curriculum content	Incorporate content about conceiving, designing, implementing and operating AI supported systems
	Update subject specific content to reflect AI use in the workplace
	Lift program learning outcomes for graduate employability
Cognitive process	Cognitive off-loading
	Free up mental space
	Higher order cognition
Pedagogical method	Virtual teachers/tutors
	Personalised/customized learning experience
	Authentic assessment
	Complete transformation questioning division in fixed set of subjects, set duration and location

).

These key areas of impact are related to how AI-supported systems primarily affect education: what is taught (Curriculum Content), how students think and learn (Cognitive Process), and how teaching and assessment are conducted (Pedagogical Method).

In terms of the curriculum content, we consider first integrating knowledge of GAI, ranging from user skills to designing and building GAI systems, depending on the specific engineering program. Because it is through engineering systems that GAI finds its practical applications, its impact on the engineering curriculum includes a necessary update to add the conception, design, implementation, and operation of GAI-enabled systems. In addition, the content of existing courses should be revised to include GAI tools, ensuring students are prepared for their future use in the workplace. Finally, the incorporation of GAI as a learning tool expands the learning outcomes and helps to develop professionals with stronger overall skills, not just in GAI but across the board, enhancing their adaptability and expertise [25].

Regarding the cognitive process, several trends are related to the possibilities of cognitive off-loading, using artificial intelligence tools to reduce the mental effort required to complete tasks. Instead of relying solely on human memory, problem solving, or decision making, individuals delegate part of the cognitive workload to GAI systems. Examples include the use of AI-powered calculators, virtual assistants, or recommendation systems. This process allows individuals to focus on higher-level thinking or creativity [34], while the GAI handles repetitive, complex, or data-intensive tasks [35]. Obviously, while this can improve efficiency and productivity, it may also lead to over-reliance on AI, potentially affecting critical thinking or problem-solving skills over time. Further research is needed to understand the impact of GAI on students’ development of these skills.

Current studies are inconclusive so far. On the one hand, the strengths and weaknesses of GAI have been studied by Farrokhnia et al. [36], who conducted a SWOT analysis for ChatGPT, as paradigm of generative GAI systems. The results show interesting strengths, such as the facilitation of personalized learning and the reduction of teachers’ workloads. However, they also identified critical weaknesses, emphasizing the importance of fostering higher-order learning outcomes, including creativity, ethical principles, and critical thinking.

On the other hand, Urbano et al. [37] applied GAI aids in educational settings and observed valuable insights into the teaching of engineering students. Their research also addressed related topics, such as designing assessment methods (as commented previously). They concluded that the use of GAI tools to support learning, contrary to the earlier concern, actually enhanced students’ critical thinking skills. This improvement was especially noticeable among master’s degree students, suggesting that advanced learners may benefit more from AI-supported education than undergraduates. These results are consistent with Narayan and Saharan (2024)’s findings [38]. Furthermore, Al Husaeni et al. (2024) [39] conducted a review of the use of smart chatbots as educational tools in science and engineering education and highlighted their use to better develop students’ 21st century skills. Empirical evidence is mounting about efficiencies when GAI is incorporated in the teaching and learning flows [40]. Based on the ability to offload and realize efficiencies, embedding GAI in teaching and learning provides an opportunity to free up space in the study plans. AI-driven platforms are being used for individual study guidance, to help students structure their assignments, and to face their exams more effectively [41]. These aspects are categorized in the taxonomy under pedagogical methods with the tags of virtual teaching and personalized learning and assessment.

However, beyond the specific contents of a study program, the adoption of GAI may lead to a much broader transformation that questions the current curriculum structure based on a fixed set of subjects of standard duration, delivered by a designated academic often at a particular location. The content is pervasive, available almost everywhere, and, with GAI, can be delivered in a way that resembles human communication [42]. This way of learning may cause a complete transformation in the organization of courses, subjects, and temporalization.

In such a transformation, considering a new paradigm where there is a cognitive offload, a possible improvement in the learning outcomes, and a new program structure of increased flexibility, there may be space to include contents that have been challenging to introduce in engineering programs and that may provide a valuable opportunity to address wider expectations in terms of sustainability, ethics, and social considerations.

7. Leveraging GAI to Integrate Sustainability Perspectives in Engineering Education

Emphasis on socio-environmental awareness, including sustainability and circular design, in the engineering curriculum is needed more than ever before, especially given the potential for GAI to accelerate and augment engineering activities and its social and environmental impact.

Previous attempts to incorporate social and environmental aspects have had limited success due to reluctance to compromise technical content, which results in tensions when trying to add sustainability aspects to the traditional curriculum [3,4,5]. This is illustrated in Figure 2 as ‘socio-technical’ tension. The adoption of GAI tools delivers teaching and learning efficiencies that create the curricular space for enhanced socio-technical integration.

Figure 2 presents a dynamic framework structured around two axes: content (technical to socio-technical) and pedagogy (lecturer control to student agency). By automating tasks and personalizing content delivery, GAI reduces cognitive load, allowing students to focus on higher-order thinking and interdisciplinary learning. This transformation supports the inclusion of ethical, social, and sustainability considerations alongside technical knowledge, bridging the gap between traditional, instructor-centered approaches and flexible, student-driven learning.

The role of engineers in society extends beyond technical problem solving; it encompasses shaping the future through the design and implementation of systems that profoundly impact people, communities, and the environment. By fully utilizing AI, the engineering curriculum can be rebalanced to equip graduates with greater command of the ethical and socio-environmental dimensions of their profession. A plausible goal would be to empower engineers to transcend the technical interest of efficiently controlling the physical world by incorporating global, temporal, and historical perspectives, as well as examining institutions and ways of thinking [43], enhancing the view of engineering as a social profession [44].

GAI can help integrate sustainability and ethics into the core of engineering programs by encouraging interdisciplinary collaboration and critical reflection. For example, students can use GAI tools to analyze data on resource consumption, assess the feasibility of circular economy models, or explore the social implications of emerging technologies [45]. These experiences not only enhance technical expertise but also cultivate a sense of responsibility and agency in shaping a sustainable future. Specifically, GAI can assist in modeling complex systems and scenarios to better understand the impact of engineering decisions on sustainability. In addition, GAI can help with analyzing large datasets, identifying patterns that can be useful for evaluating the environmental impact of large-scale engineering projects.

Furthermore, in order to ensure the responsible and ethical use of GAI in engineering, it is crucial to promote transparency and accountability. This involves competency development and the implementation of ethical principles for AI-enabled systems. Engineers must establish clear ethical guidelines addressing issues such as fairness, privacy, safety, and accountability. GAI algorithms need to be transparent and explainable. The reproducibility of the results should be also checked. Users should be able to understand how the algorithms work and how decisions are made. Engineers should also regularly evaluate the impact of GAI on the sustainability and ethics of engineering programs. This includes identifying and mitigating potential risks and ensuring that the benefits of AI solutions are distributed equitably, including the possibilities of access.

8. The Sustainability of GAI

This paper has discussed the possibility of the inclusion of GAI in curricula as a tool to further develop sustainability skills in engineering students. In doing so, it is essential to address also the sustainability of GAI and AI-enabled systems themselves. This matter is starting to attract interest in the research community [46]. One of the most significant environmental concerns of GAI is its energy consumption and associated CO₂ emissions. Training GAI models, particularly deep learning models, demands substantial computational power and data storage, leading to considerable energy and water use. Additionally, GAI impacts resource depletion and electronic waste generation. Producing hardware for GAI systems requires the extraction of metals and other materials, potentially harming ecosystems. The growing volume of electronic waste from retired GAI systems, which also occurs with similar electronic equipment, poses an increasing problem.

GAI also has significant social and economic consequences, often linked to its environmental impacts. The concentration of computational power and GAI development capabilities in large corporations can aggravate social inequalities. AI-driven automation may result in significant job losses, especially in sectors that rely heavily on repetitive tasks. These jobs are often concentrated in specific industries, such as manufacturing, retail, and customer service, which disproportionately employ individuals from certain socioeconomic groups. It is then important to explore how GAI can promote sustainable and inclusive economic development that benefits society as a whole [47]. To mitigate negative impacts and enhance the sustainability of GAI, graduate competencies are needed to develop energy-efficient GAI models using renewable energy, designing sustainable hardware, and heading toward GreenAI [48]. This can also be achieved by implementing policies and regulations by governments and international organizations, promoting GAI sustainability through incentives for sustainable GAI research, and establishing energy efficiency standards.

The next general principle to improve sustainability of GAI would be broader education and awareness on the importance of GAI sustainability and strategies to achieve it. Regarding this point, future engineers and educators play a key role. AI’s own sustainability should be a core component of engineering education. Future engineers must understand the environmental, social, and economic impacts of AI and learn to consider sustainability in their AI-enabled engineering projects.

9. Conclusions

GAI is a multifaceted tool with growing number of use cases that already cover the range of ways in which previous technological innovations have been adopted in engineering education. This involves replacing manual practices in some cases, supporting learning in others, and providing support and augmenting skills once learning has been demonstrated. Further research is needed to determine the essential competencies that must be developed without GAI’s aid and to strike the right balance between its use as scaffolding for learning and its avoidance when critical learning needs to occur without help.

Beyond the ways in which innovative technologies have been traditionally incorporated in the engineering curriculum, based on the surveyed literature, GAI has the potential to transform education more broadly by enhancing teaching methods, personalizing education, and simplifying tasks to enhance learning outcomes. Achieving the anticipated impact depends on effectively combining the expertise of educators with the capabilities of GAI tools.

GAI delivers efficiencies in teaching and learning, both for students and educators, like the increased productivity observed in other areas and in the workplace, based on automation and augmentation. This paper invites reflection on the sustainability of doing more and doing it faster, as if there was no limit. This paper also suggests a more sustainable scenario, where the efficiencies gained by introducing GAI tools are used to enhance the development of sustainability skills, including those for the design of sustainable AI systems. Given the central role of engineers in designing and implementing new AI-enabled solutions and use cases, enhanced sustainability graduate skills directly contribute to a more sustainable approach to the development of AI. As new applications of AI are explored, the question about what can be done with AI needs to expand to whether we can afford and sustain the current growth of AI-related systems.

According to the numerous references to experimenting and implementing GAI in engineering activities, a reshaping of engineering practices and, consequently, engineering education is taking place. To prepare future engineers for this transformation, the existing curricula should be revised to equip students with the skills to apply GAI solutions in a responsible and sustainable way. This preparation needs to be implemented while maintaining the basic technical foundations, critical thinking, and the main competences that form the core of engineering. The assessment of the students should be reviewed to validate that students acquire these competencies.

The use of GAI facilitates the integration of critical topics like sustainability, ethics, and social responsibility into engineering programs, areas that have traditionally been challenging to incorporate. This evolution prepares engineers to address real-world complexities with technical skills grounded in broader socio-environmental considerations.

Engineers and educators will also play a pivotal role in raising awareness and implementing sustainable practices in GAI development by addressing challenges like energy consumption, CO₂ emissions, and resource depletion.

With proper implementation, GAI has the potential to significantly enhance engineering education, equipping future professionals with the skills needed to navigate an increasingly complex and dynamic world by integrating sustainability considerations. While much of the existing literature focuses on student-centered approaches, it is equally important to consider the perspectives of educators and other stakeholders. By incorporating their insights, the efficient and effective use of GAI can be maximized, ensuring a more holistic and inclusive integration into educational practices.

The study presented has some limitations. First, the analysis is based on a scoping review of the literature rather than a systematic literature survey. However, given the evolving state of this field, we were interested in setting the background to depict a plausible future scenario, rather than conducting a comprehensive analysis to describe the current state. Second, this scoping review is based on scientific publications, not including grey literature. Therefore, although representative of the state of the art, it may not capture the whole breadth of applications of AI in engineering education. Third, our study assumes that GAI will reach its potential to provide accurate information relevant to engineering studies. Although the development of reliable GAI applications trained with validated context-specific corpora of knowledge is underway, currently, this is still a work in progress.