Data Reveals LinkedIn and X Are Flooded With AI-Generated Content
404 Media · rss · 2026-07-09
Browsing data collected via a Chrome extension by AI text detection firm Pangram reveals that mainstream social media platforms are being overwhelmed by AI-generated long-form spam. - **Platform Status**: Up to 41% of long-form posts on LinkedIn are suspected to be entirely AI-generated. On X, the combined proportion of fully AI-generated and AI-assisted long posts and articles approaches 50%. For Reddit and Substack, AI-generated content accounts for about 10% of long-form texts. - **Content Characteristics**: Long-form posts are more likely to be AI-generated than short ones. Additionally, main posts have a much higher probability of being AI-generated compared to comment replies. - **Platform Responses**: LinkedIn's built-in AI writing tools previously triggered a surge in automated content; the platform has recently tweaked its algorithm to reduce low-quality automated posts. Meanwhile, Reddit has intensified its efforts to delete AI spam, ban associated accounts, and launched an ad campaign emphasizing "human users." According to Pangram's CEO, the actual situation might be worse than the data suggests, as users installing detection plugins are inherently more sensitive to AI content.
More from AGI Musings
- Larry Fink says China is ahead in the AI energy race, citing 100 GW nuclear buildout — rohanpaul_ai · 2026-07-21
- Jamie Dimon says bureaucracy, not AI, is the real system crushing intelligence — r0ck3t23 · 2026-07-21
- OpenAI and Anthropic’s internal models are said to be far stronger than today’s public systems — scaling01 · 2026-07-21
- Superintelligence and robot abundance will force a new social contract — Dr_Singularity · 2026-07-21
- The Guardian examines how AI companionship is turning intimacy into an economy — nordicinst · 2026-07-21
- A frustrated user says modern AI keeps hallucinating on real-world repair tasks — doochenutz · 2026-07-21