More than 100 corporate accounts in South Korea using Anthropic's generative artificial intelligence (AI) service, Claude, have been reported for a potential data breach.
Concerns have arisen that account information may have been exposed due to unauthorized access to Anthropic's internal systems by an external security firm exploiting vulnerabilities in third-party software.
According to the Personal Information Protection Commission (PIPC) on November 9, Anthropic reported on October 31 that information from over 100 corporate accounts in South Korea had been leaked externally.
However, the PIPC stated that it is currently in the preliminary review stage and has not yet initiated a formal investigation.
A PIPC representative explained that investigations involving foreign companies may require a data request process prior to commencing an investigation, which could extend the review period.
The PIPC noted, "Once a formal investigation begins, we plan to conduct further inquiries to ascertain the specific extent of the damage."
Meanwhile, the breach is believed to have occurred when an external security firm accessed Anthropic's internal systems.
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.