Users will soon have a new feature that enables them to lessen the quantity of AI-created media appearing in their streams. This initiative comes as the company announced that it presently contains in excess of 1 billion content made with AI.
Trial and Implementation
The new functionality is in a pilot phase over the following weeks before a widespread launch. This development coincides with a rise in AI-generated material, fueled by emerging video-generation tools.
The Broader AI Trend
Studies have shown that almost a tenth of the fastest-growing content creators on another major platform exclusively feature AI-generated videos. A significant portion of this content is deemed mass-produced AI material, a term for poor-quality, bulk-generated content that is frequently absurd or dreamlike.
“We know from our community that numerous viewers appreciate content made with AI tools, including digital art to science explainers, and we aim to offer the community the power to see more or less of that, according to their own likes,” commented a company executive.
Scale of AI Content
The service disclosed that there are now 1.3 billion videos on its app tagged as made with AI. Overall, more than 100 m media items are submitted to the app daily, making AI content a relatively small part of its overall inventory.
Customizing Your Feed
Users will be capable of reduce or increase the amount of AI content they view by opening the topic control feature in the application and picking AI-made media. Further subjects that currently be modified under the setting include news, style, cosmetics, and dance.
Content Labeling and Policies
The platform’s guidelines require creators to tag convincing AI-generated content, and it prohibits dangerous AI impersonations of well-known individuals or critical incidents. Any authentic-looking AI video that is without a tag can be taken down under its user guidelines.
Additionally, the app will henceforth attach an artificially created watermark to content made using its own AI tools, or detected by an cross-platform project for identifying content made by AI. This step is intended to help prevent users from seeking to bypass the disclosure requirement.
Educational Initiatives
TikTok is additionally introducing a $2m AI literacy fund for professionals and institutions to develop learning resources about using AI responsibly. This initiative will support educational bodies specializing in educating programming and robotics.
Content Safety Approach
The video-sharing app has supported its content review process, following debate over intentions to make redundant a large number of content moderators in Britain. It is planning to cut hundreds of positions in its content moderation department in London, with worker representatives and internet security professionals voicing alarm that people are being substituted by AI moderation.
A senior executive clarified 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 humans. The company reported a notable drop in shocking and graphic content seen by safety staff over the previous year due to intervention by its machine learning processes.
We believe it is crucial to combine humans and technology to maintain a secure environment,” she said. “We’ll have humans as part of that process, but we additionally believe it’s extremely critical to limit contact with dangerous material as rapidly, which is the area in which a significant AI assistance is proving valuable work at speed.”
