TikTok has reportedly reduced its use of AI generated video descriptions following a wave of complaints from users who noticed bizarre and inaccurate captions appearing across the platform. The issue became widely discussed after several creators shared examples of videos being labelled with completely unrelated descriptions, including false product claims, incorrect locations and fictional events.
The automated system had originally been introduced to improve accessibility and search functions by generating summaries of video content without requiring creators to write them manually. However, many users argued that the technology was producing unreliable information that risked misleading viewers. In some cases, videos of ordinary daily activities were reportedly described as criminal incidents or breaking news stories.
Digital media analysts say the situation highlights the ongoing difficulties technology firms face when deploying generative artificial intelligence tools at scale. While AI systems are capable of processing huge amounts of content quickly, they can still struggle with context, humour and visual nuance. Critics have also raised concerns about the spread of misinformation when inaccurate AI generated text appears alongside genuine footage.
TikTok has confirmed that it is reviewing the feature and limiting where automated descriptions appear while improvements are made. The company stated that accessibility remains a priority but acknowledged that accuracy standards had not been consistently met.
The controversy arrives as social media companies face growing scrutiny from regulators over the use of artificial intelligence in consumer platforms. Industry experts believe firms will increasingly need to balance innovation with stronger safeguards to maintain public trust.
Author: Abel Vazquez Sanchez