������

UNAIR NEWS – Amid the rapid advancement of Artificial Intelligence (AI), the line between innovation and manipulation is becoming increasingly blurred. Today’s digital landscape demonstrates that AI is no longer merely a tool to assist students but has also become a technology that can expose users to misleading information and potentially harmful biases

��

In response to this growing concern, the Institute for Sustainable Community Service (LPMB) of ������ organized the “AMORA: Airlangga SDGs School X SDGs Ambassador 2026” program. The event was held on Sunday (June 14, 2026) at the Dewa Ruci Room, 2nd Floor, Airlangga Convention Center (ACC), MERR-C Campus. Serving as the keynote speaker, Ira Puspitasari, S.T., M.T., Ph.D., a lecturer at the Faculty of Science and Technology, presented a session entitled “Educational Transformation and Digital Literacy in the Era of Artificial Intelligence (AI).”

Learning from South Korea’s legal cases to everyday weather forecasts

At the beginning of her presentation, Ira encouraged participants to examine the misuse of AI technologies through the example of deepfakes. “We create innovations with the help of AI, or innovations triggered by AI. There have been many examples, one of the most widely known cases occurred in South Korea,” Puspitasari explained.

She further introduced the fundamental principles of AI, particularly the concept of machine learning. Interestingly, this technology has already become part of students’ daily lives, often without them realizing it, from weather forecasting applications to stock market analysis algorithms.

Illusion of intelligent chatbots and the risks of AI

Moving into a more in depth discussion, Puspitasa clarified common misconceptions surrounding popular chatbots such as early versions of ChatGPT. Many students assume that these platforms represent General AI capable of thinking like humans. In reality, they remain categorized as Narrow AI, designed to perform a specific task predicting the next word or token based on statistical probabilities.

“When asked to provide information that does not exist, AI will still generate the statistically most probable response. This is what causes hallucinations. The output may appear factual, but it is actually fabricated,” she warned.

One of the most critical consequences of this mechanism is the emergence of AI hallucination bias. AI systems do not possess the reasoning ability to distinguish between factual and fictional information. When prompted with questions for which reliable data are unavailable, AI will continue generating responses that appear convincing based solely on statistical patterns.

Prompting skills must go hand in hand with ethics and verification

Toward the end of her presentation, Puspitasari acknowledged that today’s younger generation, particularly ������ students, demonstrates remarkable application proficiency in using AI tools, sometimes even surpassing their lecturers in operational skills.

“Therefore, using AI requires digital literacy and sufficient knowledge. Only then can we immediately recognize whether something makes sense or is merely fabricated. We must always review the output, decide whether it is reliable, or verify it by consulting other credible sources,” Puspitasari concluded.

Author: Fauziah Laili Romadhon

Editor: Ragil Kukuh Imanto