Industry context: Financial AI practitioners must prioritize robust input sanitization and sector-specific evaluation because domain models face targeted prompt-injection, multimodal, and numeric-error risks. Reporting by The Korea Herald, Asiae, and ChosunBiz states that KakaoBank's Financial Technology Research Institute had four papers accepted at major AI conferences in 2026. Per those outlets, KakaoBank presented a prompt injection detection method at ICLR 2026 in April, two studies at LREC 2026 in May on multimodal prompt-attack detection and numeric-calculation error identification, and a joint paper with KAIST accepted to the industry track at ACL 2026. The Korea Herald reports the ICLR study used an in-house dataset of about 59,000 cases and showed stronger detection performance than prior models. "These studies are meaningful not only as academic achievements, but also as practical technologies that can improve the safety and accuracy of financial AI services," a Kakao Bank official said, per The Korea Herald.
KakaoBank lands 4 papers at leading AI conferences