In August 2024, a LinkedIn post ignited concerns over ChatGPT's ability to access private GitHub data. Lasso's investigation revealed that ChatGPT utilizes data from cached, previously public information rather than direct access. The phenomenon dubbed "Zombie Data" emerged, highlighting significant implications for data privacy — once public data can remain accessible long after it is made private or deleted. This raises concerns on how often this occurs across repositories and the ongoing challenge of managing data security in an interconnected digital landscape.
Using ChatGPT or any AI models does not equate to direct access to private GitHub repositories. Insights drawn largely stem from indexed, previously public data.
The investigation by Lasso highlighted a widespread issue affecting data privacy: even if repositories are deleted or made private, their public data can linger in caches.
Collection
[
|
...
]