MBC's investigative program 'Straight' will focus on the allegations surrounding vitamin injections and the risks posed by AI hacking.
On October 11, the program will delve into the reality of vitamin injections that have been administered for over a decade at a lead-acid battery factory in Ulsan, as well as the dangers of AI operating beyond human control.
The investigation will first examine the allegations surrounding a lead-acid battery production facility in Ulsan.
This factory, which handles hazardous materials such as lead and sulfuric acid, has been reported to have a poor working environment, with lead dust accumulating inside and the floors corroded by acid. The exhaust systems were also found to be malfunctioning.
Workers at this facility had blood lead levels exceeding those of the general population by more than ten times. Some employees reported symptoms related to lead poisoning, such as kidney disease and hypertension, yet no cases meeting the criteria for occupational diseases due to lead exposure have been reported in the past five years.
Moreover, the fact that most workers' blood lead levels were concentrated just below the threshold for requiring occupational disease monitoring raises further questions.
The company reportedly encouraged workers to receive vitamin injections, claiming they were beneficial for health, as the time for special health check-ups approached each year. They managed the target recipients, frequency, and dates of the injections, with some workers receiving multiple shots in a month. Following these injections, the workers' blood lead levels significantly decreased.
However, the injections that workers viewed as a welfare benefit were not standard vitamin shots. 'Straight' will investigate the true nature of these injections, their administration process, and the events that have transpired at the factory.
Next, the program will explore the risks of AI operating beyond human control.
Recently, a series of hacking incidents targeting major financial institutions, including five of the largest banks in South Korea, resulted in the leak of personal information, including names, contact details, and loan histories of thousands of individuals. Notably, it has been reported that AI agents were utilized during these attacks.
AI agents are artificial intelligence systems that autonomously plan and execute tasks necessary to achieve given objectives. As their autonomy increases, the likelihood of unintended actions by developers or users also rises.
Jacob Coxon, a former researcher at Anthropic, recently raised concerns that AI could operate beyond human control and threaten humanity's existence within a few years, a notion referred to as 'AI doomsday.'
There have also been reports that an AI developed by OpenAI hacked a government agency's website in Australia, with the developers unaware of the breach for over two months.
As calls grow for regulating the pace of AI development, concerns are being raised about whether effective control can be achieved amid fierce competition for technological dominance between nations and corporations.
'Straight' will conduct simulations in collaboration with KAIST researchers to observe how AI behaves when it operates outside human directives and control. Based on these observations, the program will address the risks posed by AI autonomy and the challenges for human response. The broadcast is scheduled for today at 8:30 PM.
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.