API to report abuse/inappropriate content
I suggest that the report features extend to API users. Report abuse is only available for individuals who get exposed to, click, and decide to report a tweet/account.
A lot of inappropriate content is unfortunately on twitter. Perhaps I could categorize into 4 types:
(1) inappropriate language: hate speech, harassment (bullying, sexual harassment)
(2) inappropriate image: pornography, graphic violence
(3) intention for inappropriate and harmful activities: intention to commit a crime (shooting, bomb, individual or mass threats), sell inappropriate products and services (prostitution, drugs - not debating if it is illegal or not, but I think twitter shouldn't be the platform to offer it)
(4) false information: fake news, anti-vax campaigns, fake images (generated by artificial intelligence or just simple editing)
Currently, API can report a spam account to Twitter (POST users/report_spam). Not sure if I can report multiple accounts at once or specific/multiple tweets.
My proposed solution for the report abuse API:
a. Report multiple accounts
b. Report hashtags
c. Report multiple tweets with abusive keywords
I understand there are many factors to consider what is "inappropriate". However, if you provide access for API users to report, twitter would gain more insights faster and at a larger scale to inform better decisions in taking required action and refining the community guidelines.
I agree with this. I know Twitter now has a policy / mechanism about flagging sensitive media content and accounts (both for the uploaders and general users). But it needs to be done manually, what if the uploaders don't flag their account, thus underage users have a chance to see such tweets. What if users that see those sensitive tweets don't report the tweet to be flagged as 'sensitive content'?
In addition, the policy / mechanism only applies for image, video, profile/header picture, meanwhile porn text exists too. Currently I'm working on my thesis about detecting porn tweets (text-based) in Bahasa Indonesia. It will be more useful if what is detected can be reported automatically via API. I think this is a good enough solution for preventing underage users get exposed to pornography / violence content according to Twitter's current policy.
But this far, what is flagged will only be masked or given warnings, so if the users click it, they can still access it. What I suggest is for it to be completely hidden for underage users.
We already see bad-faith, abusive, coordinated "mass reporting" of accounts in order to get them suspended.
Adding API access would only offer greater potential for such abuse.
This is, frankly, a terrible idea.
Scandal in Birdland commented
I would suggest taking point #3 out of the proposal. While all the points can (and will) be abused by the same people this intends to curb, that one in particular can be used to further antagonize sex workers.
Here's a poll that may help the case. It looks like people want to report other stuff than SPAM via the API https://twitter.com/SafeFilter/status/1175525641564954625
I agree, it would be nice to be able to report harassment, offensive images/content, etc. via the API.
Thanks for your contribution, this is very helpful! A similar idea has previously been shared on this platform, so I will merge the two suggestions.
Terence Eden commented
I second this idea. Or give us a simpler access to the report menu, right now just can't get access to it.
si dawson commented
Since Oct 2009¹, the /users/report_spam² endpoint has allowed reporting a tweet as spam (& optionally, blocking the user).
No new API reporting features have been added since 2009.
There are MUCH bigger problems Twitter is facing today than spam (at least in part due to the report_spam endpoint being available). It would be nice to be able to report some of them, too.
Obviously, replicating the entire (regularly changing) reporting UI via API would involve a fair bit of work (particularly given the variance between countries - see: reporting hate speech³ or viewing Twitter⁴ in France or Germany).
So, let's keep this simple. What would be very helpful, as a start, would be: