🔗 Share this article

TikTok is introducing a fresh option that allows them to lessen the amount of AI-made videos appearing in their streams. This move comes as the service announced that it currently hosts over 1 bn AI-generated clips.

Pilot Phase and Launch

This change is being tested over the next few weeks before a full-scale rollout. This shift aligns with a rise in content created by AI, spurred by emerging content generation software.

AI Content on Other Platforms

Studies have shown that approximately 10% of the most rapidly expanding video streams on a popular video site only show clips produced artificially. A considerable share of this material is deemed “AI slop”, a term for poor-quality, mass-produced content that is frequently nonsensical or surreal.

Our research shows that many people enjoy content made with AI tools, from AI artwork to educational videos, and we want to give people the option to see more or less of that, based on their own likes,” said a platform representative.

Scale of AI Content

The service disclosed that there are presently 1.3bn videos on its service labeled as made with AI. Collectively, in excess of 100 million media items are added to the app daily, making clips made by AI a minor portion of its overall inventory.

User Control Features

Users will be able to limit or boost the volume of artificial media they watch by using the content preference setting in the platform and selecting AI-made media. Other topics that are already be adjusted under the filter cover politics, style, cosmetics, and movement.

AI Disclosure Rules

TikTok’s policies mandate users to identify convincing AI-made content, and it prohibits harmful AI impersonations of public figures or critical incidents. Any realistic AI-generated clip that is unmarked can be taken down under its terms of service.

Furthermore, the service will henceforth add an artificially created label to content made using its proprietary AI systems, or detected by an collaborative effort for recognizing AI-generated material. This step is designed to deter users from attempting to circumvent the labeling process.

Educational Initiatives

The platform is additionally introducing a two million dollar AI literacy fund for experts and institutions to create informative material about responsible artificial intelligence. This program will assist educational bodies specializing in educating computer science.

Content Safety Approach

The video-sharing app has defended its moderation strategy, following discussion over proposals to cut jobs for numerous safety staff in the UK. The company is expected to reduce 439 jobs in its trust and safety division in London, with worker representatives and online safety experts expressing worry that people are being substituted by machine-led review.

A trust and safety lead stated that people-led review would remain essential, but artificial intelligence can safeguard staff by filtering out the most harmful content ahead of it is viewed by people. The platform reported a significant reduction in shocking and graphic content handled by content reviewers over the last 12 months due to intervention by its machine learning processes.

It is vital to integrate people and machines to maintain a secure environment,” she said. “We’ll have people as integral to the approach, but we also think it’s extremely critical to limit contact with harmful content as rapidly, which is why a significant AI assistance is proving valuable operate efficiently.”