AI-generated complaints increase insurers processing burden

Seoul Economic Daily and The Korea Times report that South Korean insurers and the Financial Supervisory Service (FSS) are experiencing increased complaint volumes and processing burdens tied to generative AI tools being used to draft submissions. Per Seoul Economic Daily, complaints filed with 39 domestic insurers in Q1 totalled 15,996, up 5.2% from Q4 and 19.3% year-over-year. The coverage attributes 11,108 complaints to non-life insurers and 4,888 to life insurers. Seoul Economic Daily reports some individuals filed as many as 40 complaints in two months.

The Korea Times reports that complaint texts often include inaccurate legal interpretations and fabricated court precedents that do not exist. One example cited: a complaint against a life insurer cited a law that, on review, did not exist. Handlers at a non-life insurer flagged another submission still containing the asterisk formatting that certain AI tools insert for emphasis, per The Korea Times. Because compliance teams must review the full content of every submission - regardless of quality - a one-page complaint padded to ten pages by AI creates proportionally higher handling load per case.

Seoul Economic Daily reports that processing time per case has risen by at least 1.5-fold. FSS data cited by Seoul Economic Daily show non-life and life insurers account for 37.6% and 11.4% of all financial-sector complaints, respectively, and that 65.4% of complaints in Q1 related to compensation and payout disputes - the category most prone to AI-assisted volume inflation.

Industry observers note that insurance already concentrates a large share of financial complaints because of contract volume and product complexity. A surge in noisy or fabricated submissions inflates caseloads, widens backlogs, and raises the marginal cost of dispute resolution without corresponding improvements in consumer protection outcomes. The Financial Services Commission has separately announced it is developing countermeasures after identifying AI-generated medical records, including diagnostic reports and itemised treatment statements, being submitted to support fraudulent benefit claims, per Insurance Business Asia.

Observers should track whether the FSS or major insurers publish guidance on automated-submission detection, whether claims-processing metrics (average handling time, backlog size) are disclosed quarterly, and whether court or regulatory records note an uptick in cases citing fabricated precedents. Vendor offerings for AI-detection, metadata-based submission filtering, and identity verification tied to complaint platforms are the likeliest near-term operational responses.