Twitter has tested a new element in its effort to limit the spread of fake and deceptive content in the app, and variable labeling system which would alert users to the latest updates and information on topics within the tweets, while providing links to relevant sources on the same.
As you can see in this example, published by a reverse engineering extraordinary system Jane Manchun Wong, the new tagging system would include three different types of inserted alerts on tagged tweets, which in many cases are triggered by certain keywords.
These labels at this stage of development are:
- “Get the latest” – This would address fast-growing news topics, providing Twitter with a way to connect users with relevant information about any tweets, including certain words, such as “choice” or “poll”, which would allow users to link to the latest reports from official sources.
- ‘Be informed‘- This could also be applied to new news, although perhaps to more sustainable news, such as COVID-19, again with links to connect people with key authorities.
- ‘Deceptive’ – The last category would be “misleading”, which would clearly indicate tweets, including suspicious claims, and provide links, again, to official sources.
The first two, as noted, could likely be triggered by keywords included in the tweets, but the latter should probably be checked by Twitter’s moderating teams before being tagged, as Twitter’s automated system could find it difficult to detect relevant context.
It could be an effective way to provide more information on evolving topics – but again, it could also overwhelm users ’timelines, depending on how many labels are applied and what actually drives their inclusion.
In this regard, it would probably boil down to Twitter’s judgment, with the platform warning only of limited, selected topics. Which could make it a sustainable, useful tool – but the days are still early and it is difficult to predict the full potential impact without the availability of all relevant details.
Twitter has confirmed that this new system is currently being tested and that it repeats the best application for new labeling capabilities. Some users have noticed that these alerts look very similar to citations, which can be confusing, while others suggest that more prominent coloring for each type of alert might be a better way to highlight these tweets in the stream.
Twitter is still looking for the best way forward, but it’s interesting to note the latest developments and how Twitter is working to develop its disinformation and warning detection process.
And before you know it – yes, this would be based on existing fact-checking processes on Twitter, which means that indeed your own team teams for moderation and third-party connections decide what qualifies for each tag. If you don’t believe Twitter can be trusted to make a real call that way, I suggest using a different platform, but Twitter works with a number of independent authorities and groups to ensure users connect on the right with information on each topic.