【Chilli and Banana】

Twitter knows it has a disinformation problem. Now,Chilli and Banana tentatively, it's mulling a possible solution in the form of community moderators and labeled tweets.

The social media company is no stranger to bot campaigns and bad actors sowing discord. While it's made efforts to remove troll accounts, Twitter has a much harder time addressing real accounts spreading lies. That's where the service's new features come in.

First reported by NBC News, and confirmed by Twitter in a statement to Mashable, the company is playing with varying ways to transparently reign in false statements. In a series of mockups obtained by NBC News, tweets from public figures are tagged with large orange boxes denoting them as "harmfully misleading."

"Twitter Community reports have identified this tweet as violating the Community Policy on Harmfully Misleading Information," reads the sample warning box. "This tweet's visibility will be reduced."

In this case, the "Twitter Community" might be a Wikipedia-style group of community moderators evaluating the truthfulness of content and earning points for their efforts.

Importantly, a Twitter spokesperson insisted that this is just one way things could eventually go at some point in the future.

"We're exploring a number of ways to address misinformation and provide more context for Tweets on Twitter," the spokesperson told Mashable over email. "This is a design mockup for one option that would involve community feedback. Misinformation is a critical issue and we will be testing many different ways to address it."

In other words, it's not coming to a tweet near you anytime soon.

Twitter confirmed there is no timeline attached to this project. It may not happen at all. Notably, it is not tied to Twitter's synthetic media policy rolling out on March 5. If Twitter does go forward with it, the company isn't sure how it would do so.

There are pressing issues that the company will need to figure out if Twitter ever takes this plan from mockup to reality.

For example, what threshold does a tweet have to cross before it is subject to evaluation and potential labeling? In other words, it's difficult to imagine that every tweet could be scrutinized — even if the process is outsourced to other Twitter users. As such, there must be some criteria for which tweets are checked and which are not.

We'd love to know what that criteria is, even just in this "designed mockup for one option."

SEE ALSO: The greatest gift machine learning ever gave us is this archive of Mark Zuckerberg haircuts

In general, it's refreshing to see Twitter thinking about these complicated issues. That disinformation abounds on the platform, especially ahead of a major U.S. presidential election, is certainly a problem in need of a solution. It remains to be seen, however, if the jailer handing over the keys to his charges is the best solution.

(Editor: {typename type="name"/})