“In the age of artificial intelligence (AI), relying solely on humans to verify misinformation has its limits. Platforms must continuously manage and disclose the biases and fairness of their algorithms,” said Chami-Young Cha, director of the Max Planck Institute for Security and Privacy (MPI-SP) in Germany and a professor at the Korea Advanced Institute of Science and Technology (KAIST), during an interview after her keynote speech at the 2026 World Korean Scientists and Engineers Conference held in Gangnam, Seoul.
Cha diagnosed that the ecosystem of fake news is changing dramatically. “In the past, misinformation spread like a patchwork, rapidly gaining traction when combined with anger,” she noted. “In the future, it will be difficult to track and respond to misinformation in this way.”
She added, “The spread of misinformation may occur over an extended period through groups of AI accounts masquerading as people.” Cha expressed concern that if AI continuously injects misinformation, it could undermine not only individual choices but also democracy. “The key issue is the intent and purpose behind the creation of AI,” she stated.
Regarding government regulations aimed at eradicating misinformation, Cha pointed out that legal and institutional measures alone are insufficient. “There are many entities responsible for fake news,” she explained, citing the creators of fake news, platforms that allow it to spread on social media, and users who share it.
She emphasized that it is practically impossible to monitor every area, suggesting that issues formed in online public discourse, such as elections and significant social topics, should be closely monitored.
Cha also assessed that technology to detect AI-generated content is still not fully developed. “While technologies to determine whether content is AI-generated are emerging, methods to circumvent these detections are also advancing,” she explained, adding that it is not yet a field that can be considered fully effective.
She pointed out that the structure of information distribution is changing. In the past, the verification of misinformation was primarily conducted by humans, but now platforms are increasingly using algorithms to manage content.
“Nowadays, many people ask AI questions before consulting others,” Cha noted. “Conversations between AI and users are not visible from the outside, making it difficult to measure what information AI provided to users and whether it was incorrect.”
As interactions between AI and users increase, Cha stressed that platforms will have a greater responsibility to verify the reliability and biases of their algorithms and explain them to society. “We need to create benchmarks for various aspects such as bias and fairness and transparently disclose the results,” she said, emphasizing that this is essential to confirm the efforts platforms are making for social value.
Cha is a data science expert researching solutions to social issues such as misinformation, poverty, and public administration using data science and AI. Her main research area is 'Data Science for Humanity.' She has conducted studies on the mechanisms of fake news dissemination using AI and big data, misinformation detection, analysis of impoverished areas based on satellite imagery, and AI for public services. Her research achievements have been recognized with the Hong Jin-ki Creative Award in 2024 and the Innovation Medal of Science and Technology this year.
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.