X Redirects Stolen-Content Revenue to Original Creators, Bypassing Traditional Takedown Model

X's upgraded Grok AI model now detects duplicated content at three times its predecessor's rate, identifying 1.5 million stolen posts as of July 2026. Rather than removing copies, X redirects monetized impressions to the original uploader — over $1 million so far. The mechanism applies to disguised video and viral text posts, shifting revenue from thieves to creators. False-positive risk at the payout stage remains the key open question.