The platform is rolling out a fresh option that enables them to lessen the amount of AI-created media appearing in their feeds. This step comes as the service announced that it presently contains over 1 bn AI videos.
Pilot Phase and Launch
This change is undergoing trials over the following weeks before a full-scale launch. This advancement coincides with a increase in content created by AI, spurred by new AI video creators.
AI Proliferation Online
Studies have shown that almost 10% of the most rapidly expanding channels on a competing service only show content made by AI. A considerable share of this output is classified as “AI slop”, a phrase for substandard, bulk-generated content that is often absurd or dreamlike.
Based on user feedback that numerous viewers enjoy content made with AI tools, ranging from computer-generated images to science explainers, and we want to give people the option to adjust the amount of that, depending on their own preferences,” stated a platform representative.
AI on the Platform
The service shared that there are now 1.3bn videos on its network labeled as artificially created. Overall, in excess of 100 m pieces of content are uploaded to the service daily, making AI content a minor part of its total content library.
User Control Features
Members will be in a position to reduce or increase the amount of AI content they see by opening the content preference feature in the application and choosing “AI-generated content”. Additional categories that currently be controlled under the filter cover politics, style, aesthetics, and choreography.
Content Labeling and Policies
Its guidelines mandate users to identify “realistic” AI-made content, and it prohibits malicious AI impersonations of celebrities or critical incidents. Any realistic AI-generated clip that is without a tag can be deleted under its terms of service.
Additionally, the platform will from now on apply an artificially created watermark to content made using its own AI tools, or detected by an industry-wide initiative for identifying artificial media. This action is designed to deter users from trying to bypass the disclosure requirement.
Educational Initiatives
The company is additionally establishing a $2m educational fund for specialists and groups to produce learning resources about responsible artificial intelligence. This project will benefit charitable organizations dedicated to training programming and robotics.
Moderation Strategy
The platform has supported its content review process, amid discussion over proposals to cut jobs for numerous review team members in the UK. TikTok is expected to cut hundreds of positions in its content moderation division in its British office, with worker representatives and online safety experts expressing worry that staff are being displaced by AI moderation.
A senior executive clarified that staff-based oversight would continue to be important, but artificial intelligence can shield workers by filtering out the most graphic content before it has to be seen by staff. TikTok reported a significant reduction in explicit and harmful content reviewed by human moderators over the previous year due to reliance on its machine learning processes.
It is vital to integrate humans and technology to keep the platform safe,” the official commented. “We’ll have humans as involved in the system, but we additionally believe it’s very important to reduce exposure to harmful content as swiftly, which is why a lot of the machine support is making a difference process content quickly.”
