AI digital forensics company Yurak said on Monday it will launch DFAS Arc, an AI-based solution that automatically classifies data by grade.

The company said DFAS Arc helps organisations respond to the transition to the National Network Security Framework (N2SF) and a data-centric security environment.

It analyses the meaning and context of data using AI and automatically classifies it into CSO (confidential, sensitive, public) grades. It also enables functions ranging from decision grounds and approval history to encryption, deletion and audit records. Rather than relying on simple keyword or pattern searches, it interprets a document's meaning and presents a grade based on relevant legal provisions and internal rules. The person in charge approves the final grade after reviewing the reasons provided by the AI. The company said this can reduce the time needed to classify large volumes of data and lower differences in judgement and the risk of misclassification among staff.

DFAS Arc can analyse not only documents but also unstructured data such as design drawings, images, video and audio, the company said. It analyses information contained in drawings or technical images to determine whether they are industrial secrets, and converts audio in meeting recordings or videos into text to check what was said, it added.

After classification, it encrypts confidential and sensitive materials based on their grade. Data whose retention period has expired or is unnecessary is deleted in an unrecoverable manner after an approval process by the person in charge. Yurak plans to expand its business first in the public sector, where demand for N2SF transition is set to take off, and then broaden applications to manufacturers holding national core technologies and industrial secrets, as well as the finance and healthcare sectors.

Yoo Bong-seok (유봉석), Yurak's chief executive, said the company differentiated the product by applying closed-network AI and integrity and traceability technologies accumulated in digital forensics to data classification. He said this enables integrated management of classification results, decision grounds and follow-up security measures. He said the company will start with the public-sector N2SF transition market and expand into areas managing important information such as industrial secrets and personal information, aiming to establish itself as a standard classification engine for a data-centric security environment.