X has announced a significant crackdown on creators who repost others' content without permission, using an updated version of its Grok AI model to detect duplicated material at three times the speed of its predecessor. The platform said it has already identified 1.5 million stolen posts in its latest detection cycle, though it did not specify the timeframe.
Under the new policy, any monetised impressions generated by stolen content — including videos, images and viral text posts — will be redirected to the original uploader. Attempts to bypass detection by adding watermarks, intros or other edits will also result in revenue being transferred away from the infringer. X's head of creator partnerships, Bier, noted that common examples of stolen text include the often-repeated line: 'Twitter is like the smoking section of the internet.'
The move comes as platforms across the social media landscape grapple with content theft, which undermines creator trust and distorts advertising revenue. X has also accelerated its bot-suspension efforts, claiming it was identifying and suspending 208 bots per minute as of April. The company now warns that repeated or deliberate attempts to circumvent the new policy — including engagement bait such as 'I'll follow everyone who replies' — will lead to removal from the creator programme and referral to the policy team for account suspension.
For UK businesses and creators, the policy shift has clear implications. Small and independent content producers who have seen their work reposted by larger accounts without credit may now receive a fairer share of advertising revenue. However, the reliance on AI detection raises questions about accuracy and the potential for false positives, particularly for parody accounts or legitimate commentary that repurposes viral material.
From a regulatory perspective, the UK Information Commissioner's Office (ICO) continues to monitor how platforms use AI for content moderation, particularly around transparency and appeals processes. Meanwhile, the EU AI Act, which came into force in stages from 2025, classifies such detection systems as limited-risk, requiring clear labelling and user notification. UK businesses operating across Europe will need to ensure compliance with both frameworks as X rolls out these changes globally.
Experts caution that while the technology is a step forward, it does not solve the underlying incentive problem. 'Detection is one part of the puzzle, but platforms also need to make it easier for creators to assert ownership from the start,' said Dr. Alice Thornton, a digital policy researcher at the University of Cambridge. 'If the appeals process is opaque or slow, smaller creators could still lose out.'