Today, Twitter unveiled a new policy that it claims would increase openness around the offensive tweets on its site that have been the target of disciplinary action. When tweets break Twitter’s rules, one of the company’s typical responses is to reduce their exposure, a process it refers to as “visibility filtering.” In these cases, the tweets are still accessible online but are less likely to be seen since they are not included in search results, trends, recommended notifications, For You and Following timelines, and other areas.

Instead, readers must go directly to the author’s profile to view the tweet.

When such enforcement occurs and advertisements won’t run against the content, the tweet may also be de-emphasized in replies, according to Twitter’s rules.

In the past, the general public might not have known whether a tweet had been modified in this manner. Currently, Twitter claims that will alter.

The business has decreased the appearance of tweets that have been flagged as possibly breaking its regulations but promises to “soon” start attaching visible labels to those posts. It did not specify a precise date for when the technology will be fully implemented throughout its network.

Additionally, the business said that not all tweets whose visibility has been decreased will be labeled.

It claims it would only initially roll out the functionality to tweets that break its Hateful Conduct code but that it will eventually expand it to other areas as well in the “coming months.”

“This change is designed to result in enforcement actions that are more proportional and transparent for everyone on our platform,” a blog post authored by “Twitter Safety” stated. The post additionally touted Twitter’s enforcement philosophy, calling it “Freedom of Speech, not Freedom of Reach.”

The business says the policy measures will happen at “a tweet level only and will not affect a user’s account.” If a tweet is labeled, the person will not be shadowbanned or removed from the network.

Twitter also explains that users whose tweets were labeled will be able to send feedback if they believe their tweet was erroneously tagged, but warns that neither it will ensure the message’s reach will be restored nor that they will receive a response.

This is probably related to the significant budget cuts Twitter made to the firm as a whole and to its Trust & Safety departments. And although it’s not obvious to what extent this system will be automated, it may significantly rely on automation to decide what should be labelled. (Since Twitter no longer responds to press queries, the only formal statements on issues like this are made in blogs and tweets by the company or Elon Musk, the new owner.) Twitter acknowledges in a Twitter conversation discussing the modifications that automation naturally increases the possibility of mistakes. Additionally, the business states that it will “in the future” permit authors to challenge its decision in this case.