SAN FRANCISCO — X is aiming its most capable artificial intelligence at one of social media's oldest problems: creators who lift other people's work and cash in on it. In an update announced on July 16, the platform said the newest version of its Grok model can now flag duplicated content at three times the rate of the previous version, and that the money earned by copied posts will be routed back to the people who made the originals. The original report on the crackdown lays out how it works.
A Smarter Watchdog
The upgrade leans on the same Grok technology that powers SpaceXAI's consumer products, now tuned to catch reposts, lightly edited clips, and copy-pasted viral text. Adding watermarks, intros, or other cosmetic tweaks meant to disguise stolen material will no longer fool the system. When Grok spots a match, the monetized impressions are reassigned to the original uploader, and the policy even covers recycled text posts that have circulated for years.
In its most recent cycle, X said it identified roughly 1.5 million stolen posts. As a result of the changes, more than $1 million in creator payouts is being handed back to the original authors rather than the accounts that copied their work. It is a concrete, dollars-and-cents win for the independent voices who make the platform worth scrolling.
Rewarding Original Work
The move fits a broader push to reward authenticity. X recently rolled out an improved video editor and recorder to nudge creators toward posting their own footage instead of reuploading someone else's, and it has leaned on Grok to suspend automated bot accounts faster than before. Repeat or intentional attempts to game the rules now carry real consequences: users caught soliciting engagement or follows three or more times can be removed from the creator program and referred for further review.
