← Front page

WSJ Tech News Briefing · Saturday, August 15, 2026

AI labels on platforms like Meta and TikTok are voluntary and imperfect

Platforms such as Meta and TikTok have implemented guidelines and tools to manage AI-generated content, including voluntary labels and visual watermarks. However, critics point out that these labels can be inaccurate, mistakenly labeling real content as AI-generated, and watermarks can be removed or edited.

companyMetacompanyTikTok

The tape

4 quotes
One way to tackle the issue is AI labels. Platforms like Meta and TikTok have put in place guidelines and created tools to manage AI content.
Nicole Nuen
But critics have noted that the labels are voluntary and the tools aren't perfect.
Nicole Nuen
A number of creators, for example, have complained that their non-AI work was mistakenly labeled as AI.
Nicole Nuen
Some companies have also begun applying visual watermarks or digital metadata to signal that a piece of content is AI. But those markers can be removed or edited.
Nicole Nuen
Heard on WSJ Tech News Briefing — “To Spot AI Fakes, Think Like a Digital Forensics Expert, published Saturday, August 15, 2026. Heardvine summarizes and quotes with attribution and timestamps, and links to the original everywhere.
Transcribed via Gemini audio transcription · $0.04
AI labels on platforms like Meta and TikTok are voluntary and imperfect — Heardvine