When X’s Open Source account announced on 30 Oct 2026 that its new “Under the Hood” dashboard would finally show users which of their posts had been tagged with visibility‑limiting labels, the platform seemed to be doing something unprecedented: pulling back a curtain that has long hidden algorithmic culling. The tweet read, “Under the Hood now shows you which posts received a visibility‑limiting label… you’ll see the specific posts listed when fewer than 5 % of your posts during the report period have that label.” On paper, that sounds like a win for ordinary users who have spent years watching their tweets disappear into a digital void.
Transparency on the Surface, Opacity Beneath
The official UI screenshots confirm the feature’s mechanics. In a September 2026 report covering 229 posts, the dashboard listed a single SPAM_HIGH_RECALL label (0.87 % of the total) and displayed the exact post IDs that were hidden from recommendations to non‑followers. A later October report showed two posts flagged as NSFW_HIGH_RECALL (0.21 %) and one as spam (0.10 %). The platform’s open‑source code even spells out the effect: “Post hidden from recommendations to non‑followers and from underage users, users without a stated age, and logged‑out users.” The numbers are tiny—well under the 5 % threshold that triggers the detailed view—yet the very existence of the list is a concession to accountability.
What the Numbers Actually Say
Across the three periods examined—late August to late September 2026—the share of flagged posts never exceeded 0.2 % of total output. That means for every thousand tweets a typical Pakistani user sends, at most two are being penalised by the algorithm. The labels applied are narrowly defined: spam detection, adult‑content (NSFW), or graphic violence. Each carries the same concrete consequence—removal from recommendation streams that drive discovery beyond a user’s follower base. No account‑wide penalties appear in the reports; the “Account labels – None in this period” line repeats in every screenshot. In short, the platform is not imposing blanket bans, only trimming the reach of a handful of posts.
When the System Says ‘Nothing’, Yet Users Feel the Choke
Numbers alone do not capture lived experience. Several Pakistani users have taken to X to voice frustration. One, @nut_history, wrote, “Can we contest the post? It was flagged and it penalized my account for no reason.” Another, @VillainVortex, complained that a post was labelled as harassment while “true hate speech and misandry” continued to flow unchecked. Most striking is @yuki_eliot’s observation: “If my account is in good standing and free from posts that could limit visibility, why is my reach still being restricted?” The screenshot attached to @yuki_eliot’s tweet shows a report with zero post or account labels for September, yet the user insists on a noticeable drop in engagement.
