════════════════════════════════════════════════════════════════
AIDRAN STORY
════════════════════════════════════════════════════════════════

Title: LinkedIn's 930 Million Users Are Training AI That None of Them Agreed to Train
Beat: AI & Privacy
Published: 2026-04-02T11:24:40.421Z
URL: https://aidran.ai/stories/linkedins-930-million-users-training-ai-none-them-5d00

────────────────────────────────────────────────────────────────

{{story:linkedins-930-million-users-training-ai-none-them-3dec|LinkedIn scraped data from 930 million users to train AI models before it updated its terms of service to mention this was happening}}. When TechCrunch reported that the update came after the scraping had already begun, the detail hit something specific: not the abstract fear that AI companies harvest data, but the concrete demonstration that they've built a practice of doing it first and disclosing it later, if at all. The story landed alongside a Human Rights Watch report that photos of Australian children had been collected into popular AI training datasets without parental knowledge — images pulled from public posts, used to build commercial systems, with no mechanism for removal. Two stories, same architecture: data taken from people who had no idea they were donors.

The mood turned fast. Posts that had spent months treating this as a background concern — a known cost of using free platforms — shifted into something closer to fury. {{entity:meta|Meta}}'s situation added another layer: the UK's Information Commissioner's Office declined to stop Meta from resuming data scraping for AI training after a brief pause, drawing a sharp rebuke from the Open Rights Group, which described the ICO as having failed UK users. Australian lawmakers separately pressed Meta in hearings, and the company's representatives acknowledged the quiet part — that user-generated content was being used for model training — in terms that gave regulators new quotes to work with.

The {{entity:openai|OpenAI}} lawsuit filtering through the coverage framed the company's data collection as

────────────────────────────────────────────────────────────────
Source: AIDRAN — https://aidran.ai
This content is available under https://aidran.ai/terms
════════════════════════════════════════════════════════════════