X is rolling out a new in-app video editor and recorder on iOS, a release tied directly to comments from head of product Nikita Bier acknowledging that a substantial share of viral video on the platform isn't original. On July 6, 2026, Bier posted that many videos from top-performing accounts are lifted from other users — in some cases reposted five years after they first went viral — and noted that video now accounts for close to half of all impressions on X The Verge.

The new toolset includes trimming, automatic caption generation, multi-language caption overlays, and a green-screen feature that lets creators composite custom backgrounds from either camera roll photos or existing posts The Verge. The stated intent, per Bier, is to lower the friction of producing original video directly in-app, giving creators fewer reasons to lean on reposted clips scraped from elsewhere.

The scale of the problem Bier described is notable given how central video has become to the platform's engagement metrics. If video drives roughly half of impressions, the integrity of that content pool has direct bearing on ad load, creator payouts, and the revenue-share program X uses to court creators away from YouTube Shorts, TikTok, and Instagram Reels. Bier's framing suggests the company sees recycled content less as a moderation nuisance and more as a structural threat to that economic model.

This is not X's first intervention this year. In May 2026, Bier said the team had identified large accounts programmatically reuploading content specifically to game the revenue-share payout system, and that X would begin reallocating impressions from those posts back to the original creator The Verge. A month earlier, in April, Bier announced X was cutting payouts to aggregator accounts that profited from stolen reposts and clickbait framing The Verge. Detection has also drawn direct engineering investment: replying to a user whose video had been stolen, Bier said the incident "really bothered us" and that X was "now investing a significant amount of engineering resources into these models" built to identify duplicated video content Nikita Bier.

Taken together, the sequence reads as a three-part response: cut off the payout incentive for reposting, build detection models to reassign credit and impressions to original creators, then lower the cost of producing original content so the incentive to repost weakens further. Bier has stated plainly that creators who avoid recycled content will climb faster on the platform than those who don't The Verge, which functions as both a policy statement and a signal to the creator economy about where the algorithm's weight is shifting.

The mechanics of "programmatic reuploading" Bier references are familiar to anyone who has watched aggregator networks operate on any platform with a payout structure: scrape trending clips, strip metadata, repost with minimal alteration, and let volume do the work. What's distinct here is the multi-year decay problem Bier flagged — content resurfacing five years after its original virality — which points to a detection gap that simple hash-matching or duplicate-frame comparison likely can't fully close, given re-encoding, cropping, and watermark removal are trivial for reuploaders to perform at scale.

Worth flagging: shipping a first-party editor addresses the supply side of the problem by making native creation easier, but it doesn't by itself solve detection of already-stolen content circulating from accounts with no intention of ever using the new tools. The engineering investment in detection models Bier mentioned is the more consequential lever, and its effectiveness will be difficult for outside observers to verify absent published false-positive/false-negative rates or transparency reporting.

The broader context here is a platform trying to compete on short-form video against incumbents with years of head start on native creation tooling, while simultaneously trying to clean up the incentive structure that made recycled content profitable in the first place. Those two goals can be in tension: better editing tools increase video volume, which increases the surface area detection systems have to cover. Whether the revenue-reallocation mechanism scales cleanly, or whether it simply pushes reuploaders toward more sophisticated obfuscation, is the question worth watching over the next several quarters rather than the next release cycle.