YouTube claims this is an accident [0] which I find a little hard to believe.
Is it crazy to think this is CCP state actors? [1] Both in the form of teams of people or bots reporting anything that they dislike to trigger Google's automation, or even just getting people hired by these companies to work on the inside for their interests.
My guess is that someone pro CCP sneaked the keywords which trigger the deletion into a list of Chinese insults (i.e. on a list of purely offensive words).
At lest one of the words found first can indeed be seen as denouncing descriptions of communism. I can't judge if it's just that or quite insulting.
If this is true then the interesting question is how they sneaked it in there. From outside by social enginering? From inside by affiliated devs? Through an consultant company hired to create a list of Chinese insults?
The most ironic think is that if my guess is true then it might literally have sneaked in without anyone intention by just using a list of insults from somewhere else without cross validating it.
My guess is that it's the CCP abusing the flagging system to trick the existing anti-abuse systems into doing the dirty work for them... An infiltrator would be harder to get in place and much easier to find and deal with (if they did something as crass as manipulate a word list).
Great! Let's fix it! Suppose you just throw the words on a whitelist? Well, now there are a couple magic words you can put in unrelated spam comments to ensure that they don't get deleted.
Honestly, all the people knotting their underwear about this don't seem like they've ever seriously thought about abuse on large platforms. It's an inherently adversarial environment, and you have to game out second- and third-order consequences for pretty much everything. And even then there will be unintended consequences. A 'real' fix that doesn't break lots of other things takes time.
The "communist bandits" thing is a Taiwan vs PRC thing -- Mao was basically a bandit for like 10 years while fighting the ROC government. Or the revolutionary vanguard expropriating wealth for the people's liberation, if you'd prefer.
Seems roughly equivalent to calling someone "you syrian terrorist" to me? Not sure where to calibrate it.
That context doesn't change my comment. Affiliation with the CCP, past or present, is an ideological affiliation. 'Syrian' is not. Comparable to 'Syrian bandit' would be 'Chinese bandit'. If you want a Syrian analogue to 'communist bandit', you might try 'ba'athist bandit'.
'Communist' and 'ba'athist' are ideological affiliations, as are affiliations with the Chinese Communist Party or the Arab Socialist Ba'ath Party specifically.
'Syrian' and 'Chinese' are not ideological affiliations.
It could be an action from a rogue employee. Possibly not a management decision or specific company policy, just a filter added by an employee with access to it. There might be some 'lively' discussion internally in Google right now over this specific issue. In the past some Google employees have taken actions against individuals based on their personal political beliefs or other personal grudges. A good example of that was when Jordan Peterson was temporarily locked out of all Google services, including his Gmail, insiders said it sparked a lot of internal arguments at the company (eventually his accounts were reinstated with no official explanation or apology, though some employees did apologize to him privately for their coworkers behavior).
It's likely at its core a problem with poor employee screening, insufficient training and supervision, and vague/over-reaching policies given to employees that they sometimes interpret as legitimizing them to censor or ban based on their personal political beliefs.
Some employee there likely is a communist sympathizer, or has other connections to China's authoritarian ruling party.
I used the term “lesbian” in a comment on Facebook this week in a respectful and topical manner.
I received a notice that the comment was removed for violating community standards against “hate speech” and was told that’s my one warning for the next 12 months and any subsequent infringements would lead to a 24 hour ban. I used their process to disagree with the finding and they replied in a few minutes that they rejected my appeal.
And now you're on double secret probation as one who has not only engaged in "hate speech", but had the temerity to protest their notice that you have been convicted and found guilty of hate speech.
Seems more believable than China infiltrating a foreign company to block two words from appearing in Chinese on a website blocked in China.