‘보이지 않는 AI’의 두 얼굴… 범죄 잡는 좋은 AI, 사람도 잡는 나쁜 AI

A recent crime in Seoul's Gangdong District involved a 30-year-old man, Kim G, who followed and pushed a 50-year-old woman, Park N, causing her injury. Kim claimed he acted without intent, but authorities used AI video analysis and other legal scientific methods to prove he intentionally pushed Park. The use of AI to catch criminals has been a topic since the 1990s, popularized by the 2002 film Minority Report.

The 2002 film depicts a high-tech security system called "PreCrime" that predicts and prevents crimes before they occur. However, this fictional setup still portrays AI as capable of predicting and moving on its own. The first practical application of AI in law enforcement occurred about 10 years ago in South Korea. Since then, it has been widely adopted by financial authorities, national forensic research institutes, and cybersecurity companies to detect repetitive criminal behaviors and unusual activities.

Today, AI is an inseparable part of crime investigation and is used for various purposes such as identifying deepfakes and plagiarism. The article points out the dangers of relying too heavily on AI's judgment, citing cases where automated systems led to wrongful accusations. The author emphasizes the importance of human judgment and the need to critically evaluate AI technology to prevent similar mistakes in the future.