Short answer is yes. Long answer is that with text it’s much easier to stamp out illegal activity because keyword searches are cheap while semantic searches in images are pretty good but extremely computationally expensive. You can’t just scan for illegal activity in images the same way you can nigh instantly scan a body of text for “illegal-site.com”.
Wow, really? I guess that unfortunately makes sense. I have a dock for my work laptop that charges and works for HDMI/etc but it uses an entire two USB-C ports at once.