The dismantling of prosecutorial authority and reorganization of indictment bodies creates an enforcement vacuum that AI-based automated prosecution systems and private justice platforms rush to fill.
The reformers wanted to check prosecutorial overreach. What they got was a prosecutorial vacuum. The new Independent Indictment Agency, stripped of the old prosecutors' institutional memory and investigative networks, cannot keep pace with its caseload. Backlogs grow. White-collar criminals walk free on statute-of-limitations technicalities. The public loses faith in the system. Into this gap steps a new class of legal technology: AI systems trained on decades of case law that can assemble evidence chains, identify applicable statutes, and generate prosecution-ready filings with minimal human oversight. Some are public utilities funded by the government; others are private platforms funded by victims' subscriptions. The justice system does not collapse — it transforms into something its architects never imagined, where the line between public prosecution and private litigation dissolves.
It is June 2032 in Seoul. A 37-year-old fraud victim sits in her apartment, staring at her laptop screen. The prosecution of her case by the Independent Indictment Agency stalled fourteen months ago. Last week, she subscribed to JustiChain Premium for 89,000 won per month. The AI has already assembled a more comprehensive evidence package than the human prosecutor managed in a year — cross-referencing bank records, surveillance footage timestamps, and witness statements into a coherent narrative. She clicks 'submit' and watches the filing upload to the Agency's portal. She feels relief, then a slow dread: justice now comes with a monthly subscription, and she wonders what happens to victims who cannot afford one.
AI prosecution tools could democratize access to justice rather than stratify it. If governments adopt these systems as public infrastructure — free, transparent, and auditable — they could eliminate the class bias inherent in human prosecution, where wealthy defendants hire better lawyers than under-resourced prosecutors can match. Algorithmic prosecution, properly regulated, might be more consistent, less susceptible to political pressure, and faster than any human system. The key variable is not the technology but the governance model: public utility versus private platform determines whether AI prosecution empowers citizens or exploits them.