iWorld
LinkedIn tops social media platforms for AI-generated long-form posts
Pangram study says over 40 per cent of LinkedIn long-form posts are AI-written
MUMBAI: Looks like the keyboard has found a new co-worker and it doesn’t need coffee breaks. A new study suggests artificial intelligence is no longer just helping professionals polish their posts on LinkedIn; in many cases, it’s writing them almost entirely.
According to research by AI detection startup Pangram, more than 40 per cent of long-form posts on LinkedIn are fully AI-generated, making the professional networking platform the most AI-saturated among major social media platforms analysed in the study.
LinkedIn also accounted for nearly two-thirds of all AI-generated posts identified during Pangram’s analysis, underscoring the platform’s growing reliance on generative AI for professional content creation.
The report found that AI-written content has become commonplace across social media, although the prevalence varies by platform and content length. Overall, the average AI-generated content rate stood at 13.8 per cent across all scanned posts.
Longer posts were significantly more likely to be AI-generated. Pangram found that one in four long-form posts defined as content exceeding 250 words was entirely created by AI, with the rate reaching 25.72 per cent across all platforms studied.
The trend was particularly pronounced on LinkedIn, while X also recorded high levels of AI-assisted writing, with nearly half of its long-form articles containing AI-generated text.
Pangram noted that on four of the five platforms examined, longer content was consistently more likely to be AI-generated than shorter posts, suggesting users increasingly rely on AI tools for drafting detailed articles, thought leadership pieces and professional commentary.
The findings highlight how generative AI is rapidly reshaping online communication, particularly on platforms where users are expected to publish polished, insightful and professionally relevant content. As AI writing tools become more accessible, distinguishing between human-written and machine-generated content is emerging as a growing challenge for both readers and platforms alike.
