Gadgets
August 14, 2026
1 views
2 min read

X is testing a tool that will let users see if their posts have been 'shadowbanned'

Curated by Patrick
Source: Engadget
X is testing a tool that will let users see if their posts have been 'shadowbanned'
Tech Daily Byte Analysis

X announced that a small cohort of testers can now access a new settings page where they can request a data dump of their account’s recent activity. The exported JSON contains month‑long aggregates of “visibility‑impacting labels” such as NSFW, spam, violence, and a “civic integrity” flag that confines a post to the author’s profile. The rollout coincides with the company’s latest GitHub release that adds dozens of content‑type tags, including a niche “emergency use only” label. X’s VP of Product Keith Coleman framed the move as “unprecedented transparency,” but the tool does not break down performance on a per‑tweet basis, requiring users to sift through raw code to spot any restrictions.

The experiment builds on X’s incremental open‑source strategy that began with earlier versions of its “for‑you” ranking code. Unlike those prior releases, which offered little practical insight, the new repository lists concrete moderation tags, suggesting a willingness to expose the mechanics that govern reach. This aligns with Elon Musk’s public pledge to make every line of X’s code auditable by third parties, a promise that has faced criticism for withholding ad‑related and brand‑safety modules. Competitors such as Meta and TikTok have been more guarded about algorithmic disclosures, so X’s partial openness could pressure the industry toward greater accountability, even if the current dataset remains coarse‑grained.

The immediate effect is that power users can now substantiate claims of “shadowbanning” by pointing to specific labels attached to their content. However, the aggregate nature of the data and its JSON format limit usability for the average user, and the absence of per‑post metrics may still leave many disputes unresolved. Moreover, exposing label definitions could enable malicious actors to reverse‑engineer the system and tailor content to evade restrictions. Watch for X’s next update, which may add brand‑safety tags or expand tester access, and for any public findings from the third‑party experts who pre‑reviewed the code. The degree to which X will eventually release the ad‑ranking and recommendation logic will be a key barometer of Musk’s transparency commitment.

Key Takeaways

X’s new “under the hood” page lets a select group download aggregate visibility‑label data for the past month.

The released label taxonomy includes NSFW, spam, violence, civic‑integrity, and an “emergency use only” flag, but no explicit political tags.

The tool provides only aggregated JSON output, making per‑post verification difficult for most users.

Future expansions could broaden label coverage or open more of the recommendation stack

About the Source

This analysis is based on reporting by Engadget. Here is a short excerpt for context:

The feature is rolling out as the company open-sources more of its ranking algorithm.
Read the original at Engadget

More in Gadgets