I agree with this. We can see the skew from that graph but it depends what data points we look at. Daily users probably skew younger than monthly or total users. I also agree that I don't think the median would get as low as 16 years but even having half of your users under 18 isn't a good sign.

That's why I said the claim in the first place is bullshit. Obviously TikTok will pick the best data points to make their argument. They have all of the data and can cheery pick something to make their case.

Ironically, this is probably far less relevant than the algorithms that they use to push content to people. For example, anyone under 14 in China can't use TikTok for more than 40 minutes per day but that restriction doesn't exist here. And those users are also shown different content (educational) instead of the regular content. So the algorithm in China for youth is different from the algorithm for youth here.