Twitter Tests New Sensitive Content Appeals Process as it Continues to improve its Discovery Efforts

This would be good news for Twitter users who feel they have been unfairly punished by Twitter’s visual identification algorithms for not perceiving the content.

Today, Twitter is releasing another test of an updated appeals process for classified media violations to make it easier to reference such statements.

The process provides a more direct way to appeal to tweet content, allowing users to get a response from Twitter if their content was accidentally tagged.

Twitter’s image recognition processes

It has been the cause of several high-profile misclassification incidents on this front. Last month, the platform was attacked after an image of a badly burned army veteran saluting in uniform on July 4 was flagged for including “potentially sensitive content”.

Realistically, Twitter needs to identify AI as much as possible to limit the spread of malicious content, and its systems are driving improvements in this regard.

In its most recent transparency report, It reported that tweet violations – those that violate rules for any reason – accounted for less than 0.1% of all tweet impressions over the period, with its tools. Artificial intelligence helps to significantly reduce exposure over time. The officials also reported a 322% increase in sensitive media policy violations and a 194% increase in accounts caused by consensual nudity.

In general, these results are good, but more activity and reliance on automatic and preventative restrictions will also lead to more false positives.

So this improved appeals process makes sense and, hopefully, will allow Twitter to optimize its systems to follow while serving users who have been inadvertently sanctioned.

A small group of users in Mexico, the United States, and Japan will make the first features of the new features available, and the last two are biggest user markets.

“If you’re participating in the experiment and you haven’t marked your media as confidential in the settings, you can dispute a tweet marked by the tweet itself (web only) or a message marked as a favorite.”

This is a great update, which is tied to Twitter’s broader shift towards better user protection.


Also Read:

Translate ยป