K Bank Launches "Smishing Text AI Detection" Service

K Bank announced on January 22 that it has launched a "Smishing Text Artificial Intelligence (AI) Detection" service.

The "Smishing Text AI Detection" service uses AI technology within the K Bank app to determine whether a text message poses a fraud risk. If a user receives a message suspected of being a scam, they can copy the message and paste it into the "Financial Safety" section under the "Full Menu" in the K Bank app.

If the AI determines that the message has a high risk of fraud, it provides a warning stating "There is a high possibility this is smishing," along with a detailed explanation of the reasoning. For example, it identifies specific scam tactics such as "posing as a work-from-home job" or "malicious links disguised as delivery confirmations." The service also provides recommendations such as ▲not clicking on links ▲being cautious when entering personal information ▲directly verifying with the relevant institution.

The core of this service is a finance-specialized private large language model (LLM), which was first introduced by an internet bank in February last year. The AI has been trained on real smishing cases and financial fraud data to analyze the words and types included in text messages, providing the likelihood of smishing and the basis for its assessment.

In particular, a machine learning system has been established that enables the AI to predict scam types and learn autonomously. The AI generates and pre-learns virtual smishing messages that may emerge in the future. This enables the detection system to immediately respond to increasingly sophisticated new smishing tactics.

A K Bank representative stated, "We will continue to introduce innovative security services based on AI technology so that customers can use our services with peace of mind."