X has launched an improved version of its Grok AI algorithm to combat content theft throughout the network, eliminating roughly 4,000 creator revenue-sharing accounts and flagging many of them for potential suspension. Nikita Bier, X’s head of product, made the announcement via a platform post on July 16, 2026. Bier stated that the most recent version of Grok detects duplicated content three times faster than its previous model and that the system can now identify reposted videos and copied text posts even when users try to disguise them with watermarks, introductions, or other edits such as content modifications that were previously enough to fool the detection system.
Instead of rewarding the account that reposted the content, X said monetised impressions from duplicated posts will now be credited to the original creator. The platform also said it detected 1.5 million stolen posts during its latest review cycle, though it did not specify the time period covered. As a direct result of the changes, more than $1 million in creator payouts will be redirected to the original owners of the content instead of the accounts that copied it.
The crackdown covers videos, images, and viral text posts. Bier cited the most common example of copied text: a widely circulated phrase that reads “Twitter is like the smoking section of the internet” — proof that basic text copying at scale has been a persistent and lucrative form of content theft on the platform. Repeated or intentional attempts to circumvent the detection system will now result in removal from the revenue programme.
“Some updates on the creator rev share program: Our new model now detects duplicated content at 3x the rate of the previous model. Adding watermarks, intros and other edits will send monetized impressions to the original uploader. We detected 1.5 million posts that were stolen this cycle. With these changes, over $1 million will be given back to original content creators.”~Nikita Bier
Engagement Baiters Are Also Out: Grok Catches All of Them Now
The update covers two distinct categories of creator programme abuse. The first is content theft – reposting someone else’s original video, image, or text and collecting the ad revenue it generates. The second is engagement baiting – posts that explicitly solicit artificial engagement, such as “I’ll follow everyone who replies” or “like this for good luck.” Bier said that accounts that solicit engagements three or more times will be removed from the creator revenue-sharing programme and forwarded to the policy team for suspension review.
He added that Grok now catches all instances of this kind of behaviour. The nearly 4,000 accounts removed on July 16 were largely identified through this engagement bait detection sweep, with the stolen content enforcement running in parallel. Both actions are part of what X has described as a broader push to clean up its monetisation ecosystem and reward creators who actually produce original work. The platform has paid out more than $100 million to creators in 2026 alone under its revenue-sharing programme.
“X will use Grok AI to better detect stolen content, redirect payouts to original creators, and crack down on engagement bait. The newest Grok model detects duplicated content at 3x the previous rate. Over $1 million in creator payouts will now go back to the rightful owners.”~TechCrunch
Grok’s Role Goes Beyond Chatbot: X Is Using It as a Platform Enforcement Tool
The use of Grok to fight content theft and engagement bait is a significant shift in how X uses its in-house AI. While the public has primarily focused on Grok as a conversational AI assistant integrated into the X app, the platform is increasingly using it as a backend enforcement layer, detecting fraud, identifying policy violations, and protecting creators’ financial interests without requiring human moderators to review posts on a large scale.
Bier noted that in April 2026, X had already been identifying and suspending 208 bots per minute using its automated systems, and that the trend is rising. The content theft detection update represents a further expansion of that AI-driven enforcement infrastructure. For creators who produce original content, the practical impact is straightforward: less revenue will be lost to accounts that copy viral posts, and the detection system is now robust enough that disguising stolen content no longer works as easily as it once did.
“X is using its Grok AI system to crack down on content theft and engagement baiting. The new Grok model is 3x better at detecting duplicated content. Adding watermarks or edits to stolen content will now redirect monetised impressions to the original creator. 4,000 accounts were removed from the creator programme today.”~Social Media Today
What It Means for X’s Creator Economy and Who Is Most at Risk?
X declared 2026 the Year of the Creator and has been pushing hard to position itself as a viable platform for original content monetisation. The revenue-sharing programme, launched under Elon Musk’s ownership, pays eligible creators a share of ad revenue generated from impressions on their posts. The crackdown on content theft is as much about the credibility of that programme as it is about fairness to individual creators. If the platform’s highest-earning accounts are simply reposting other people’s videos, original creators will have little financial incentive to create new content on X, weakening the entire creator economy pitch. The accounts most vulnerable to the new Grok detection are those that rely primarily on reposting viral videos, screenshots of popular content from other platforms, or verbatim copies of well-known text articles.
For those accounts, the combination of three-times faster detection and automatic impression redirection effectively eliminates the revenue model they have been relying on. For genuine creators, the changes represent a meaningful improvement at least on paper though the platform’s broader financial health, its advertiser relationships, and the overall scale of its creator payouts relative to rivals like YouTube remain important factors in whether X can genuinely compete as a creator platform.
“X is using Grok to hunt down content thieves and redirect creator earnings to the rightful owners. The updated Grok model detects stolen posts 3x faster, can see through watermarks and edits, and has already identified 1.5 million stolen posts — with $1 million in payouts being redirected.”~Firstpost
