The platform is rolling out a updated setting that allows them to lessen the amount of AI-created videos appearing in their feeds. This move comes as the service announced that it currently hosts over 1 bn AI videos.
Pilot Phase and Launch
This change is in a pilot phase over the next few weeks before a global release. This shift aligns with a surge in AI content, fueled by recent AI video creators.
AI Content on Other Platforms
Recent reports that approximately one in 10 of the quickest rising channels on a competing service exclusively feature clips produced artificially. A considerable share of this output is deemed low-quality AI content, a term for poor-quality, factory-style content that is typically illogical or surreal.
Based on user feedback that numerous viewers like content produced via AI tools, from digital art to science explainers, and we aim to give people the ability to see more or less of that, depending on their own likes,” said a company executive.
Scale of AI Content
The platform shared that there are now 1.3bn videos on its network identified as artificially created. In total, over 100 mn media items are added to the service each day, making AI-generated material a minor percentage of its overall inventory.
Customizing Your Feed
Members will be in a position to reduce or increase the level of artificial media they watch by opening the “manage topic” feature in the application and picking “AI-generated content”. Further subjects that currently be modified under the setting include news, fashion, aesthetics, and dance.
AI Disclosure Rules
The platform’s policies oblige users to identify convincing AI-generated content, and it bans dangerous fake videos of well-known individuals or critical incidents. Any believable AI video that is unmarked can be deleted under its community policies.
Furthermore, the platform will henceforth apply an “AI-made” label to content made using its own AI tools, or identified by an collaborative effort for identifying content made by AI. This step is designed to help prevent users from seeking to evade the disclosure requirement.
Educational Initiatives
The platform is additionally launching a $2m educational fund for specialists and organizations to produce learning resources about using AI responsibly. This project will benefit charitable organizations dedicated to training coding and technology.
Content Safety Approach
The platform has supported its content review process, following discussion over intentions to make redundant hundreds of review team members in the United Kingdom. It is set to reduce 439 jobs in its trust and safety department in its British office, with trade unions and digital protection specialists voicing alarm that humans are being replaced by machine-led review.
A trust and safety lead stated that people-led review would continue to be important, but AI can protect employees by eliminating the most harmful content ahead of it is viewed by people. TikTok reported a notable drop in shocking and graphic content handled by content reviewers over the previous year due to reliance on its machine learning processes.
We believe it is crucial to combine people and machines to maintain a secure environment,” the executive stated. “We’ll have humans as involved in the system, but we also think it’s very important to limit contact with unsafe media as quickly as possible, which is where a lot of the machine support is making a difference operate efficiently.”
