Today, there was an update to the X algorithm.
https://x.com/XOpenSource/status/2087951962004230428
Today we take another big step in our ongoing commitment to transparency. We are open-sourcing the code that powers the For You timeline and releasing a new tool that displays labels applied to accounts and posts that may limit their visibility.
As stated, you can access this new tool, "Under the hood," at the link below:
https://x.com/i/under_the_hood
It is currently in a pilot testing phase. To download the report, you must have "at least 10 posts in the previous month" and "an account that is over one year old." (Note: Since it is a test period, it may only be available to a limited number of users.)
When I downloaded it, the file looked like this:
startDate: 2026-07-01
endDate: 2026-07-31
timezone: UTC
→ The target period is July 1st to July 31st (last month).
generatedAt: 2026-08-08T23:59:59Z
→ The date and time this report was generated.
postCount: 456
→ 456 posts during the target period.
postLabels: [ ]
→ No labels applied at the post level.
accountLabels: [ ]
→ No labels applied at the account level.
totalPostLabels: 0
→ Total of 0 labels on posts.
totalAccountLabels: 0
→ Total of 0 labels on the account.
If there were applicable labels, it seems they would be listed here with their counts.
It appears to be a feature that allows you to check whether your posts or account are subject to restrictions.
X also posted the following:
https://x.com/kcoleman/status/2087952961045414211
"Am I shadowbanned?" "Is X fair?" "Why am I seeing this post?"
We want the public to be able to answer these questions for themselves. Today, we are providing an unprecedented level of transparency into the X algorithm, allowing people to audit, critique, and even help improve it. Enjoy & send us your feedback.
Until now, even if you felt like you were being "shadowbanned," it was difficult for users to accurately confirm that status.
By being able to personally check the restriction labels actually applied to your account or posts, X's transparency has arguably reached a new level.
Of course, just because there are no labels doesn't mean your posts will definitely go viral, but being able to distinguish between "restricted and not spreading" versus "simply having a low algorithmic evaluation" is a major change.
As this is a pilot test period, the specifications may change, but I have summarized the information known so far!
I will share more updates as soon as they become available.
