The Korea Insurance Research Institute announced in a recent report that the insurance industry should rearrange the rules on consumer rights and data privacy protections to prepare for broader use of artificial intelligence (AI) tools.

The latest study emphasized the need to address liability concerns related to AI, particularly in cases of AI-driven incomplete sales. The study recommended regulatory measures and guidelines for use of AI tools in order to tackle issues and safeguard consumer rights.

The paper added that insurers in the United States have already been grappling with challenges following AI adoption, citing instances where some health insurance claims were denied with minimal human review, often as short as 1.2 seconds per case.

The Personal Information Protection Commission, a national data protection authority, in the meantime, has issued a new guideline on pseudonymization of data such as voice, video, and text, all used to develop AI technology.

The guideline specified that using unstructured data in autonomous driving systems and traffic data sharing requires operators to blur faces and license plates to prevent computer identification.

For AI chatbots, it mandated a clear separation between the “learning database” and “answer database” to avoid revealing pseudonym information used for language learning as responses.

The guideline also introduced some revisions.

Under the guideline, for example, how appropriate and reliable processing technology works should be reviewed by an independent committee of experts.

Organizations using pseudonym information must enhance internal controls to prevent personal information breaches and ensure that pseudonym data is discarded after use.

By Yoo Joon-ho, Lee Dong-in, and Han Yubin

[ⓒ Pulse by Maeil Business News Korea & mk.co.kr, All rights reserved]