ENDLESS: International Journal of Futures Studies Vol.
8 No.
Education Beyond AI: Building Integrity Through Authentic Assessment Wahidin1.
Soegianto Soelistiono2.
Rona Tanjung3 Univeritas Siliwangi.
Tasikmalaya.
Indonesia Universitas Airlangga.
Surabaya.
Indonesia Universitas Riau.
Riau.
Indonesia Email: wahidin@unsil.
Abstract The advancement of artificial intelligence (AI) has transformed the paradigm of higher education assessment, challenging the effectiveness of traditional examination methods such as essays and multiple-choice tests.
AI technologies, like ChatGPT, enable students to generate high-quality responses rapidly, triggering an authenticity crisis in academic assessments.
Research indicates that 56% of global students, including those in Indonesia, use AI for assignments or exams, with 54% considering it a form of cheating (BestColleges, 2.
This paper aims to analyze the weaknesses of traditional exams in the AI era and propose authentic evaluation approaches through a conceptual discussion based on the analysis of challenges and new strategies.
Key findings include recommendations for project-based exams, assessments of learning processes, hands-on practical exams, and the integration of AI literacy and ethics into the curriculum.
This transformation demands that lecturers serve as facilitators and authentic evaluators while developing curricula focused on competence, creativity, and soft skills.
Consequently, higher education can produce graduates who are not only academically proficient but also critical, creative, and responsible in leveraging AI.
Keywords: Artificial Intelligence.
Higher Education.
Traditional Exams.
Authentic Assessment.
Project-Based Exams.
Process-Based Assessment.
AI Literacy.
AI Ethics.
Academic Integrity.
Curriculum Transformation.
INTRODUCTION
The rapid advancement of artificial intelligence (AI) technologies, such as ChatGPT.
Claude.
Copilot, and Gemini, has significantly transformed the landscape of higher education.
These technologies enable students to access, process, and present information quickly and efficiently, opening new opportunities for learning, such as personalized educational experiences and enhanced efficiency (Holmes et al.
Rizki et al.
, 2.
However, the ease of AI use also poses serious challenges to traditional evaluation methods, such as essay exams, multiple-choice tests, and written assignments, which have long served as the standard for assessing student competence (Cotton et al.
, 2.
Research indicates that 56% of global students, including those in Indonesia, have used AI to complete assignments or exams, with 54% considering such use as cheating or plagiarism (BestColleges, 2.
An authenticity crisis emerges as students can produce high-quality responses in seconds without deep understanding, as evidenced by a surge in AI-related academic misconduct cases in UK universities, reaching 7,000 cases in the 2023-2024 academic year (The Guardian, 2.
In Indonesia, the potential for cheating is heightened as students increasingly rely on AI for written tasks, bypassing http://endless-journal.
com/index.
php/endless ENDLESS: International Journal of Futures Studies Vol.
8 No.
independent effort (Lancaster, 2023.
Hidayat & Santoso, 2.
This suggests that traditional exams are no longer adequate for measuring student competence in the AI era (Perkins & Roe, 2.
The challenge is compounded by the lack of specific policies in many Indonesian universities to regulate AI use, despite plans to integrate AI into the national curriculum starting in 2025-2026 (Universitas Gadjah Mada, 2024.
Susanti & Nugroho, 2.
This study aims to analyze the weaknesses of traditional exam methods in the AI era, propose new authentic evaluation approachesAisuch as project-based exams, process-oriented assessments, and hands-on practical examsAiand discuss the role of lecturers as facilitators and authentic evaluators.
Additionally, it explores the transformation of higher education systems, including adaptive curricula and an emphasis on AI literacy and ethics (Fitria, 2023.
Ali et al.
, 2.
The scope of this research focuses on higher education, particularly student evaluation, considering both global and local contexts in Indonesia, where AI adoption is growing but faces challenges like digital disparities and data privacy concerns (Helmiatin et al.
, 2024.
The Atlantic, 2.
Thus, this study contributes to the discourse on how higher education can adapt to produce graduates who are competent, creative, and responsible in the AI era (Crawford et al.
, 2023.
Times Higher Education, 2024.
Stanford Graduate School of Education, 2023.
The Guardian, 2024.
Inside Higher Ed.
Statista, 2.
LITERATURE REVIEW
Traditional examination methods, such as essays, multiple-choice tests, and written assignments, have long been the cornerstone of assessing student competence in higher education.
Essay exams measure studentsAo ability to formulate written arguments, multiple-choice tests evaluate memorization and basic conceptual understanding, while written assignments assess research and information synthesis skills (Cotton et al.
, 2.
However, advancements in artificial intelligence (AI) technologies, such as ChatGPT.
Claude, and Copilot, have exposed significant weaknesses in these methods, particularly regarding the authenticity of assessments (Perkins & Roe, 2024.
Holmes et al.
, 2.
One of the primary weaknesses is the ease with which AI can complete traditional exams without requiring deep understanding from students.
Research shows that 56% of global students, including those in Indonesia, use AI to complete assignments or exams, with tools like ChatGPT capable of producing 2,000-word essays in seconds or answering memorization-based multiple-choice questions with high accuracy (Best Colleges, 2023.
Statista, 2.
For instance, in Indonesia, students increasingly rely on AI for written tasks, bypassing independent research processes, which heightens the risk of academic misconduct (Lancaster, 2023.
Hidayat & Santoso.
Globally.
AI-related cheating cases in UK universities have surged, with 7,000 students detected using AI in the 2023-2024 academic year, rising from 1.
6 cases per 1,000 students in 2022-2023 to 5.
1 per 1,000 (The Guardian, 2.
In Australia, pressure http://endless-journal.
com/index.
php/endless ENDLESS: International Journal of Futures Studies Vol.
8 No.
to pass students suspected of using AI further threatens academic integrity (The Guardian, 2.
This vulnerability has triggered an authenticity crisis, where grades no longer accurately reflect studentsAo abilities or understanding.
Studies indicate that 54% of global students view using AI for assignments or exams as cheating, yet its use remains widespread due to unclear regulations (Best Colleges, 2023.
Inside Higher Ed.
In Indonesia, despite efforts to integrate AI into curricula, specific policies to prevent AI-related cheating in universities remain limited, exacerbating threats to academic integrity (Universitas Gadjah Mada, 2024.
Susanti & Nugroho, 2.
Traditional approaches focusing on memorization and final products are susceptible to AI manipulation, failing to capture studentsAo critical thinking and creativity (Fitria.
Ali et al.
, 2.
Consequently, traditional evaluation systems are no longer adequate for measuring student competence in the AI era, necessitating a transformation toward more authentic assessment approaches (Crawford et al.
, 2023.
Rizki et al.
, 2025.
The Atlantic, 2024.
Times Higher Education, 2024.
Stanford Graduate School of Education, 2023.
Helmiatin et al.
, 2.
METHOD
This study adopts a conceptual approach grounded in the theoretical framework of authentic learning theory and the perspective of technological ethics in education (Holmes et al.
, 2023.
Crawford et al.
, 2.
An analytical approach is employed through a literature synthesis to identify the limitations of traditional examinations in the AI era and to formulate authentic evaluation solutions, such as project-based assessments and process-oriented evaluations.
These solutions are supported by empirical data indicating that 56% of global students use AI for academic tasks or exams (BestColleges, 2.
and reported cases of academic misconduct in UK universities (The Guardian, 2.
The framework integrates concepts of AI literacy and digital ethics to address challenges related to academic integrity, particularly in Indonesia, where AI regulations remain limited (Rizki et al.
Susanti & Nugroho, 2.
The analysis also considers global and local contexts by comparing assessment practices and educational policies, ensuring relevance to the dynamics of higher education (Perkins & Roe, 2024.
Universitas Gadjah Mada, 2024.
The Atlantic, 2024.
Times Higher Education, 2.
Table 1 compares traditional examinations with authentic evaluations based on five key aspects: authenticity, resistance to AI manipulation, creativity, student engagement, and academic integrity.
Traditional examinations, such as essays, multiple-choice tests, and written assignments, score low in authenticity and academic integrity due to their vulnerability to AI-assisted cheating, with 56% of global students reportedly using AI for tasks or exams (BestColleges, 2.
The increase in academic misconduct cases, such as 7,000 instances in UK universities during 2023-2024, underscores these weaknesses (The Guardian, 2.
In contrast, authentic evaluations, including project-based assessments, process-oriented evaluations, and hands-on practices, are rated highly for fostering critical thinking, http://endless-journal.
com/index.
php/endless ENDLESS: International Journal of Futures Studies Vol.
8 No.
collaboration, and resistance to AI manipulation (Holmes et al.
, 2023.
Rizki et al.
In Indonesia, authentic assessment approaches are beginning to be adopted, though challenges such as the digital divide hinder implementation (Universitas Gadjah Mada, 2.
This table highlights the need for a transformation in evaluation practices to ensure assessments that accurately reflect studentsAo capabilities in the AI era (Perkins & Roe, 2024.
Hidayat & Santoso, 2.
Table 1: Comparison of Traditional Exams and Authentic Evaluations Assessment Aspect Traditional Exams Authentic Evaluations Authenticity Low High Resistance to AI Manipulation Very Low High Creativity Moderate High Student Engagement Moderate High Academic Integrity Low High Caption: Qualitative scale (Very Low.
Low.
Moderate.
Hig.
based on literature analysis.
Source:
Adapted from BestColleges .
The Guardian .
Holmes et al.
, and Rizki et al.
RESULTS AND DISCUSSION
Alternative Evaluation Approaches To address the authenticity crisis in higher education assessments triggered by artificial intelligence (AI), traditional evaluation methods must be reformed to adopt more authentic and technology-resistant approaches.
Research indicates that AI tools, such as ChatGPT, can rapidly produce high-quality responses, undermining the validity of essay-based and multiple-choice exams (BestColleges, 2023.
Cotton et al.
Consequently, alternative evaluation approaches that prioritize creativity, critical thinking processes, and ethical literacy are essential to ensure assessments reflect studentsAo true abilities (Perkins & Roe, 2024.
Rizki et al.
, 2.
This section outlines four key approaches: project-based and collaborative exams, process-oriented assessments, hands-on practical exams and presentations, and the integration of AI ethics and literacy.
Project-Based and Collaborative Exams Project-based exams emphasize real-world problem-solving that demands synthesis, creativity, and collaborationAiskills that AI struggles to replicate (Crawford et al.
, 2.
This approach directs students to apply knowledge in practical contexts, such as physics students tasked with creating a simulation of physical phenomena using programming code and explaining the underlying logic, or literature students writing reflective essays based on personal experiences connected to literary themes (Holmes et al.
, 2.
In Indonesia, this approach is being adopted by some universities to foster critical thinking and originality, though challenges like limited infrastructure persist (Helmiatin et al.
, 2024.
Fitria, 2.
This method aligns with global recommendations to diversify evaluation metrics, reducing reliance on textbased outputs vulnerable to AI manipulation (The Atlantic, 2.
http://endless-journal.
com/index.
php/endless ENDLESS: International Journal of Futures Studies Vol.
8 No.
Process-Oriented Assessment Process-oriented assessment focuses on evaluating the stages of learning rather than solely the final product, capturing the dynamics of studentsAo cognitive processes.
Components assessed include class discussions, document revision logs, weekly reflections, and task revisions, enabling lecturers to gain insight into studentsAo thought processes (Stanford Graduate School of Education, 2023.
Hidayat & Santoso, 2.
The goal is to ensure that assessments reflect studentsAo effort and understanding, rather than outputs potentially generated by AI.
In Indonesia, this approach is particularly relevant for addressing studentsAo tendencies to use AI to bypass independent work, as highlighted in studies on contract cheating (Lancaster, 2.
Globally, process-oriented assessments have proven effective in enhancing academic integrity amid rising AI-related cheating (The Guardian, 2024.
Inside Higher Ed.
Hands-On Practical Exams and Presentations with Digital Tool Support Hands-on practical exams, such as oral presentations, open-ended questionand-answer sessions, or field-based tasks, allow lecturers to assess studentsAo understanding in real time, minimizing opportunities for AI manipulation (Times Higher Education, 2.
The integration of digital tools like YouTube enables students to record presentations, fostering confidence in on-camera performance and facilitating asynchronous assessment (Fitria, 2.
Moreover, recorded presentations uploaded to platforms like YouTube can be shared as learning resources, allowing other institutions to observe and adopt innovative classroom dynamics (Susanti & Nugroho, 2.
In Indonesia, this approach is being adopted in some academic programs, though challenges such as limited technology and lecturer training persist (Universitas Gadjah Mada, 2.
Globally, universities in the UK report that oral exams and field-based tasks have reduced AI-related cheating, with 7,000 cases detected in the 2023-2024 academic year (The Guardian, 2.
This approach not only strengthens academic integrity but also enhances studentsAo communication skills and confidence (Crawford et al.
, 2.
Integration of AI Ethics and Literacy The integration of AI ethics and literacy into assessments aims to train students to use AI responsibly, rather than as a tool for plagiarism.
Exams may include tasks requiring students to explain their process of using AI, including modifications made to AI-generated outputs and the rationale behind them (Ali et al.
, 2023.
Statista, 2.
The objective is to foster ethical AI literacy, which is critical in Indonesia, where specific policies on AI use remain limited (Rizki et al.
, 2.
Globally.
AI literacy training has been recommended to address concerns about data privacy and cheating, with 79% of students receiving AI ethics discussions from lecturers (BestColleges.
The Atlantic, 2.
This approach supports the development of soft skills and digital ethics, which are essential for graduates in the AI era (Crawford et al.
, 2023.
Times Higher Education, 2.
http://endless-journal.
com/index.
php/endless ENDLESS: International Journal of Futures Studies Vol.
8 No.
The Role of Lecturers in the AI Era The advancement of artificial intelligence (AI) has transformed the dynamics of higher education, demanding a more adaptive role for lecturers to maintain academic integrity and ensure the relevance of learning (Holmes et al.
, 2023.
Rizki et , 2.
With 56% of global students, including those in Indonesia, using AI for assignments or exams, lecturers are no longer solely evaluators but also curators, facilitators, assessors of original thought, and mentors for ethical AI use (BestColleges.
This role is crucial in addressing the authenticity crisis caused by AIAos ability to produce high-quality responses without deep understanding, as evidenced by the rise of 7,000 AI-related cheating cases in UK universities during the 2023-2024 academic year (The Guardian, 2.
This section outlines three primary roles for lecturers in the context of higher education in the AI era.
Curator and Facilitator As curators and facilitators, lecturers are responsible for designing learning experiences based on real-world problems that challenge students to think critically and creatively, such as scientific experiments, field observations, or social projects (Crawford et al.
, 2023.
Fitria, 2.
This approach ensures that assignments cannot be completed with a single AI command but instead require synthesis and collaboration (Helmiatin et al.
, 2.
In Indonesia, universities such as Universitas Gadjah Mada have begun adopting project-based approaches to enhance student engagement, although these efforts are limited by infrastructure and faculty training (Universitas Gadjah Mada, 2024.
Susanti & Nugroho, 2.
Globally, this approach has been proven effective in developing student skills that are difficult for AI to replicate (The Atlantic, 2024.
Stanford Graduate School of Education, 2.
Evaluator of Original Thought As evaluators of original thought, lecturers are tasked with assessing the originality of student work through face-to-face discussions.
Q&A sessions, and stepby-step reviews that allow for the identification of individual thinking processes (Perkins & Roe, 2024.
Hidayat & Santoso, 2.
This approach is effective in distinguishing student work from AI-generated outputs, particularly amid rising instances of academic dishonesty.
For example, in Australia, pressure to pass students using AI threatens academic integrity (The Guardian, 2.
In Indonesia, face-to-face discussions and step-by-step reviews are increasingly used to combat contract cheating, which shares similarities with AI-related misconduct (Lancaster, 2.
This approach also supports the development of critical thinking skills, a global priority in assessment reform (Times Higher Education, 2024.
Inside Higher Ed, 2.
Mentor for Ethical AI Use As mentors, lecturers have the responsibility to guide students in understanding the limitations and potential of AI in a responsible manner, fostering digital literacy and ethics (Ali et al.
, 2023.
Statista, 2.
With 79% of global students http://endless-journal.
com/index.
php/endless ENDLESS: International Journal of Futures Studies Vol.
8 No.
receiving discussions on AI ethics from their lecturers, such training is crucial to prevent plagiarism and promote transparency in AI use (BestColleges, 2.
Indonesia, where specific policies on AI use are still limited, lecturers can integrate assignments that require students to explain their use of AI and how they have modified its outputs (Rizki et al.
, 2025.
Susanti & Nugroho, 2.
This approach aligns with global recommendations to build ethical awareness and protect data privacy, preparing students for a technology-dominated workforce (The Atlantic, 2024.
Times Higher Education, 2.
Transformation of the Education System The presence of artificial intelligence (AI) demands a fundamental transformation of the higher education system to maintain relevance and academic integrity amid technological challenges (Holmes et al.
, 2023.
Rizki et al.
, 2.
With 56% of global students, including those in Indonesia, using AI for assignments or exams, traditional approaches that focus on content delivery and memorization-based tests are no longer sufficient (BestColleges, 2.
Educational transformation is necessary to produce graduates who are competent, creative, and responsible in the AI era, especially given the rise in academic misconduct, such as the 7,000 cases reported in UK universities during 2023Ae2024 (The Guardian, 2.
This section outlines three key elements of transformation: adaptive and contextual curricula, portfolio-based assessments, and emphasis on soft skills and meta-learning.
Adaptive and Contextual Curriculum Higher education curricula must shift from content delivery to the development of critical thinking, creativity, and collaboration skills (Crawford et al.
Fitria, 2.
This approach emphasizes learning that is relevant to real-world challenges, such as solving social problems or applying technology in local contexts (Helmiatin et al.
, 2.
In Indonesia, plans to integrate AI and coding into the curriculum starting in 2025Ae2026 reflect a step toward adaptive curricula, although challenges such as the digital divide remain obstacles (Universitas Gadjah Mada, 2024.
Susanti & Nugroho, 2.
Globally, adaptive curricula have been proven to enhance student skills that are difficult for AI to replicate, such as contextual analysis and innovation (The Atlantic, 2024.
Stanford Graduate School of Education, 2.
Portfolio-Based Assessment Portfolio-based assessment offers a holistic picture of student capabilities through collections of projects, video presentations, and reflective recordings, reducing reliance on one-time exams that are vulnerable to AI manipulation (Perkins & Roe, 2024.
Hidayat & Santoso, 2.
This approach enables evaluation of both the process and outcomes of learning, such as project revisions or critical reflections, which showcase student development (Lancaster, 2.
In Indonesia, portfolio assessment is beginning to be adopted in several universities, although limited technological infrastructure remains a challenge (Rizki et al.
, 2.
Globally, http://endless-journal.
com/index.
php/endless ENDLESS: International Journal of Futures Studies Vol.
8 No.
universities in the UK and Australia report that portfolios are effective in reducing AIrelated cheating and enhancing academic integrity (The Guardian, 2024.
Inside Higher Ed, 2025.
Times Higher Education, 2.
Emphasis on Soft Skills and Meta-Learning The development of soft skills and meta-learningAisuch as self-directed learning, critical thinking, and digital ethicsAihas become crucial in the AI era, where information is easily accessible through technology (Ali et al.
, 2023.
Statista, 2.
This training prepares students for a dynamic job market, with 79% of global students receiving discussions on AI ethics from their lecturers (BestColleges, 2.
Indonesia, the emphasis on digital ethics is especially relevant given the limited regulation on AI use in universities (Susanti & Nugroho, 2.
Globally, the focus on meta-learning has been recommended to foster adaptive skills that support lifelong learning (The Atlantic, 2024.
Times Higher Education, 2.
Research Ethics and Policy The ethical use of artificial intelligence (AI) in Indonesian universities is governed through official guidelines from the Ministry of Education.
Culture.
Research, and Technology (Kemendikbudriste.
, particularly the Guidelines for the Use of Generative Artificial Intelligence in Higher Education (Kemendikbudristek, 2.
These guidelines outline five principles of trustAifairness, transparency, resilience, clarity, and data privacyAias well as ten principles for generative AI use, including accountability, literacy, and collaboration (Kemendikbudristek, 2024.
Rizki et al.
Universities in Indonesia are encouraged to integrate AI into the curriculum with clear usage policies, including sanctions for ethical violations, ranging from written warnings to suspension for serious offenses such as data breaches (Susanti & Nugroho, 2.
However, many institutions still lack formal policies, causing confusion among academics (Lancaster, 2023.
B et al.
, 2.
National regulations such as Law No.
27 of 2022 on Personal Data Protection and the Minister of Communication and Information Technology Circular No.
9 of 2023 reinforce data protection and AI ethics, although implementation at the university level remains hindered by the digital divide and limited training (Universitas Gadjah Mada, 2024.
The Atlantic.
This approach aligns with global recommendations for AI literacy and clear ethical policies (Holmes et al.
, 2023.
Times Higher Education, 2.
CONCLUSION
The emergence of artificial intelligence (AI) has transformed the landscape of higher education, demanding a fundamental shift from traditional evaluation methods toward authentic approaches focused on project-based exams, processoriented assessments, and hands-on practice (BestColleges, 2023.
Cotton et al.
, 2.
With 56% of global studentsAiincluding those in IndonesiaAiusing AI for assignments or exams, conventional methods such as essays and multiple-choice questions have become vulnerable to manipulation, sparking an authenticity crisis, as reflected in http://endless-journal.
com/index.
php/endless ENDLESS: International Journal of Futures Studies Vol.
8 No.
7,000 cheating cases in UK universities during 2023Ae2024 (The Guardian, 2025.
Perkins & Roe, 2.
Authentic approachesAisuch as project simulations, process reflections, and oral presentations supported by platforms like YouTubeAihelp reduce reliance on AI outputs and foster creativity and critical thinking (Holmes et al.
, 2023.
Hidayat & Santoso, 2.
The implications of this transformation are significant for both lecturers and the educational system.
Lecturers must shift from being mere assessors to becoming facilitators and mentors, designing real-world problem-based learning experiences and guiding students in the ethical use of AI (Crawford et al.
, 2023.
Ali et al.
, 2.
An adaptive curriculum that emphasizes competencies, collaboration, and soft skillsAisuch as digital ethics and meta-learningAiis necessary to prepare students for a technology-dominated workforce (Fitria, 2023.
Rizki et al.
, 2.
In Indonesia, challenges such as the digital divide and the lack of AI regulation must be addressed to support this transformation (Universitas Gadjah Mada, 2024.
Susanti & Nugroho.
Therefore, higher education must prioritize the development of graduates who are creative, critical, and responsible through reforms in assessment and curriculum.
Recommendations include adopting portfolio-based assessments, implementing AI literacy training, and establishing clear ethical policies to prevent misconduct, as supported by both global and local practices (The Atlantic, 2024.
Times Higher Education, 2024.
Stanford Graduate School of Education, 2023.
Inside Higher Ed, 2025.
Statista, 2024.
Lancaster, 2.
Through these steps, higher education can produce graduates who are ready to contribute meaningfully in the AI era.
REFERENCES