Bot is a very vague word, covering many different things.

Crawlers do not try to hide that they are bots, simply looking at the user agent is enough to tell they are a bot.

Social media bots try to hide that they are bots, both to the social media and to services like cloudflare. Any half-decent social media bot will exit through a domestic IP, not through an hosting provider IP, especially not in Yandex's AS. And their traffic will be categorized as human.

This graph is likely from a site proxy (e.g. Cloudflare, but I think the colours are wrong) showing information about incoming requests. Posting bots generally make much less traffic than scrapers (scrapers tend to blindly follow links, while posting bots just make the API requests they want).