Mostly True
The claim that AI Hub discloses 150 types of public data and that personal information within them is being re-examined has been judged largely accurate, with one caveat: the datasets in question are a subset of the service's scope, making the "150 types" figure dependent on how the terms are defined.
AI Hub, the public AI data platform operated under the Ministry of Science and ICT, lists 150 types of open data available to developers and researchers. The figure reflects the platform's public data catalog as currently presented on its site.
Alongside the catalog, attention has turned to whether the datasets contain personal information and how it is handled. The renewed review focuses on privacy content within the publicly released collections, an issue that recurs whenever large-scale data sets compiled from real-world sources are redistributed for AI training.
The "150 types" description carries an expression-level caveat: the data in question forms a subset of the service's overall targets rather than its full scope. For that reason, the claim is assessed as broadly correct rather than exact in every respect — the number itself stands, but it does not describe everything the platform offers.
Verdict: Mostly True