Google uses non-human automation to make some decisions, including banning accounts. As others have mentioned, this is not unreasonable as long as there is a reasonable (in terms of time and effort) path to disputing a ban - i.e., speaking to a human about the issue. But Google (and Facebook, and probably some other companies) don't have reasonable processes for disputing or resolving these situations. Some have said…
In a setting where advertisers are effectively forced to use Google to avoid giving market share to competitors, there's the element of not having a choice while ending up with a significant disadvantage once these mechanisms falsely trigger.
With Google being the operator of the platform and judge at the same time, I don't think they can hide behind terms of use in all jurisdictions. Scaling up without carrying the costs involved seems pretty unjustified.