Academic Assessment and fair use of AI

It's been two decades of working in academia. Teaching, learning, and instructional planning at higher education institutions and its assessment are multidimensional processes. In universities, learning is typically assessed through both closed-book and open-book exams, for evaluating different dimensions of a student's knowledge and skills. Closed-book exams test a student's ability to recall and apply knowledge from memory, emphasizing mastery of core concepts, critical thinking, and problem-solving under time constraints. These exams are designed to assess how well students can retain information and think critically without external resources. In contrast, open-book exams focus on evaluating a student's ability to locate, analyze, and synthesize information effectively. They encourage deeper understanding and application of knowledge, as students are often required to interpret complex problems, make connections across topics, and justify their reasoning using external materials. While closed-book exams measure memory and conceptual clarity, open-book exams emphasized different capacities like research skills, deeper comprehension, and the ability to apply knowledge in theoretical scenarios or problem solving.

The advent of search engines and online libraries significantly transformed open-book exams, shifting the focus from merely accessing information to critically analyzing and critiquing or using it further. Previously, open-book exams often relied on physical materials like textbooks and class notes, requiring students to spend considerable time locating relevant content and writing. However, with vast digital resources at their fingertips, students could quickly search for specific concepts, access diverse perspectives, and retrieve up-to-date information in real-time. This change pushed educators to design more complex and application-based questions that go beyond simple retrieval of facts. Open-book exams now often emphasize higher-order thinking skills, such as evaluating sources, integrating interdisciplinary knowledge, and solving real-world problems. The shift also underscored the importance of digital literacy, as students must discern credible sources from unreliable ones and effectively navigate the wealth of available online information.

The use of AI differs from traditional search engines or digital libraries in its ability to provide more personalized, and interactive support. Search engines and digital libraries primarily function as tools for retrieving information from a vast database of sources based on user queries. Users must go through search results or library catalogs to locate, evaluate, and synthesize the information they needed. In contrast, AI can go beyond simple retrieval by understanding the user's intent, summarizing information, generating insights, or even creating new content tailored to specific needs. For example, while a search engine might return a list of articles about climate change, an AI tool can synthesize key points from those articles, explain complex concepts, or generate arguments for a debate. AI is like personalized tutors, providing real-time feedback, answering follow-up questions, and adapting to a learner’s style or level of understanding. This makes AI not just a tool for accessing information but also a partner in learning and problem-solving, enhancing productivity and creativity in ways traditional search engines or libraries cannot.

Using AI in exams is often considered cheating even in open book exams and assignments if it is not referred and acknowledged. This is because AI can undermine the integrity of the assessment by providing students with answers, solutions, or content that they may not have derived on their own. Unlike traditional tools like search engines or calculators, which require users to actively search, analyze, and synthesize information, AI can generate complete responses, solve complex problems, or even write essays with minimal input. This blurs the line between a student’s independent work and external assistance, potentially giving an unfair advantage. Exams or assignments are typically designed to evaluate a student’s understanding, critical thinking, and ability to apply knowledge. When AI is used to bypass these cognitive processes, it compromises the authenticity of the assessment and makes it difficult to gauge the student’s actual abilities. Furthermore, the use of AI can violate academic integrity policies if it involves unauthorized assistance, misrepresentation of work, or plagiarism. For these reasons, many institutions are debating how to integrate AI responsibly in education without undermining the purpose of assessments.

Here the nstitutions and instructors must establish clear policies, rules, and regulations to ensure the fair and ethical use of AI in academic settings. These policies should define acceptable and unacceptable uses of AI, outlining when and how students may use AI tools in learning and assessments while emphasizing academic integrity. Transparent guidelines can help students understand their responsibilities and prevent misuse, such as using AI to complete exams or assignments dishonestly. At the same time, instructors need to develop more robust and innovative assessment methods that go beyond evaluating rote memorization or formulaic answers. By designing tasks that prioritize critical thinking, creativity, and the application of knowledge in unique contexts, instructors can create assessments that are less susceptible to AI-generated shortcuts. Combining clear policies with thoughtfully designed assessments will allow institutions to leverage AI as a tool for enhanced learning while maintaining the integrity and fairness of academic evaluations.