Their methodology (comparing hashtag counts) does not make sense for detecting whether content is systematically promoted and demoted to fit some agenda. I can make one video, have it artificially promoted to reach a million viewers, and that's a hashtag increment of 1. Or I can make ten videos that are all unfairly demoted and therefore barely get any views, and that's a hashtag increment of 10.

The discrepancy for Kashmir-related hashtags makes me think there are a lot of Pakistanis on TikTok who never use Instagram...

Glad that I never installed that.

I have heard about these claims since tic-tok's inception. Glad there is actual data to back it up.

This says nothing about data capture through the app: it is solely about artificially promoting or suppressing of content within the app.