February 13, 2026 | 10:58 am
The use of AI is potentially disrupted by various forms of bias. Human-generated data is never entirely fair.
RELYING entirely on artificial intelligence, or AI, for decision-making is like looking into a cracked mirror. This technology carries an inherent imprint from its creators in the form of bias, arising from algorithms trained on imperfect human historical data. Rather than being objective, AI-driven decisions risk perpetuating prejudices that cause harm.
AI bias is nothing new. Amazon Inc. detected it more than a decade ago when the company tested an automated employee recruitment program. The experimental program turned out to be gender-biased, granting preference to men over women. The cause was the résumé data submitted to the company, which was dominated by male applicants. As a result, the system “learned” that male candidates were more desirable.
Biased data is the most common source of AI bias. If the data used to train the AI’s brain—known as large language models (LLMs)—reflects historical inequalities, whether based on gender, race, class, or geography, the output will also be biased. For example, if data on world leaders is dominated by men, AI will come to associate leadership with men.
Another source of bias comes from human design choices and assumptions that determine what is considered important, which criteria and variables are used, and which are ignored. If designers are unaware of their own biases, AI will simply replicate them. Bias can also originate from the end-users, the so-called usage bias. This can occur when users merely seek justification for preconceived views.
Therefore, one way to reduce AI bias is to encourage users to ask critical questions, which is not to accept AI answers at face value, but to test them, challenge them, and examine the assumptions behind them. For instance, when ChatGPT generates a list of world leaders that is masculine, militaristic, and Western-centric in response to the question “who is the greatest leader of all time,” a critical user would ask, “why are there no women on this list?” By doing so, the bias is exposed rather than remaining hidden.
In the Indonesian context, the issue of AI bias is becoming increasingly relevant alongside the rapid spread and use of generative AI. A 2024 Amazon Web Services survey of over 2,000 business leaders and members of the public showed a steadily rising level of AI adoption. Approximately 18 million business units—equivalent to 28 percent of all companies—have adopted AI. Around 10 to 11 percent of companies use AI for strategic decision-making.
As the number of AI users increases significantly, AI literacy in Indonesia remains low. According to a survey by the Indonesian Internet Service Providers Association, the AI Literacy Index stands at just 49.96 percent, indicating an insufficient level of understanding to optimize this technology. In other words, many users are able to use AI to complete tasks, but are not yet skilled at distinguishing AI-generated decisions from human judgment.
AI literacy has become an urgent necessity because AI is no longer merely a technological experiment, but a force that is reshaping how people learn, work, and interact. The World Economic Forum’s 2025 report estimates that AI will affect nearly all industries and shift skill demands in the global labor market. Without AI literacy, we are vulnerable to becoming users who merely consume technology without understanding its limitations, ethics, and potential risks.
Why Indonesia's Unpad Rector Allows AI Use in the Classroom
1 hari lalu
The use of artificial intelligence at Unpad must be regulated within the framework of scientific integrity, according to the rector.
India Tightens AI Rules for Social Media Platforms
1 hari lalu
Social media platforms will have to clearly label AI content and comply with takedown requests by Indian authorities within three hours.
University of Indonesia Rector Explains New AI Program
10 hari lalu
The University of Indonesia (UI) believes it is essential to prepare human resources equipped with specific competencies in the AI field.
Google Gemini Arena Makes Its Debut at Gunadarma University
21 hari lalu
To strengthen AI literacy among students and academics, Gunadarma University became the first campus to present Google Gemini Arena at a university.
Indonesia Temporarily Blocks X's Grok AI Feature
33 hari lalu
Indonesia's Komdigi has temporarily blocked X's Grok AI feature, citing risks linked to harmful and pornographic AI-generated content.
Indonesian Police: Grok AI Photo Manipulation Could Be Criminally Punishable
35 hari lalu
The National Police's Criminal Investigation Department through the Cyber Crime Directorate is currently investigating the use of photo manipulation in applications such as Grok.
Grok AI Under Fire as Users Exploit Platform to Sexualize Photos
38 hari lalu
The Grok AI tool has been misused to digitally remove clothing from people in photos, including minors.
Deepfakes Running Rampant, Here's How to Spot Fake AI-Generated Faces
45 hari lalu
More cases show that the general public is more likely to perceive AI-generated faces as original compared to real human faces.
Google's Parent Alphabet Secures $4.75bn AI Power Supply
50 hari lalu
A number of tech giants, including Google, are increasingly developing self-sufficient energy supplies to power their data centers.
Meta to Launch 'Mango' and 'Avocado' AI Models in 2026
52 hari lalu
This project is managed under Meta's superintelligence laboratory led by Alexandr Wang, a founder of Scale AI.
