With the development of AI-based tools, content creation has never been easier. But in the face of “deepfakes” and other content generated by artificial intelligence, certain social networks have decided to take action. TikTok allows you to flag this type of content using a new tool.
• Also read: The author of “Game of Thrones” and other authors file a lawsuit against OpenAI
• Also read: Amazon expands Alexa with generative artificial intelligence
• Also read: TO SEE | The artist’s stunning geometric works created with AI are captivating social media
Tom Cruise’s “deepfakes” caused a stir on TikTok at the beginning of 2021. A deepfake is a “hyperfake” that allows a person’s voice and physique to be recreated using artificial intelligence. This often stunningly realistic content can cause confusion on social networks if users do not understand that it is not the real actor or actress. In fact, the creators of this content are not obliged to inform users about the use of AI in their content.
That is no longer the case. TikTok just launched a new label that will label videos generated using artificial intelligence. Content creators must apply it to “any content generated entirely by AI or significantly modified,” as TikTok explains in the official announcement. This new feature also allows creators to comply with the Synthetic Media Policy introduced in March last year. “This policy requires that AI-generated content that contains realistic images, audio or video content be labeled to help viewers contextualize the video and prevent the potential spread of misleading content.” Creators can do this now do it using the new label (or other type of information, such as a sticker or caption). If a creator does not follow this new rule, they risk having their content deleted from the application.
But even though the platform leaves it up to creators to add this tag themselves, the Chinese giant has still been working on a way to automatically detect this content. “This week we will begin testing an ‘AI-generated’ label that we plan to automatically apply to content that we determine has been edited or created using AI. To further clarify TikTok’s AI-powered products, we are also renaming TikTok AI effects to specifically include “AI” in their name and corresponding effect label, and we have updated our Effect House developer guidelines to to do the same,” explains the social network.
The Meta Group is also working on a feature that warns that content has been generated by AI. And other platforms should also follow the same path.
These new features are due not only to the social networks’ strategy to combat deepfakes, but also to the platforms’ commitment to respect the Digital Services Act. Since August 25, this European regulation on digital services requires the identification of such AI-generated content for reasons of transparency.
The European Parliament also passed a law last June aimed at monitoring and regulating the use of artificial intelligence. The AI law is expected to be finalized by the end of the year following discussions and negotiations between Parliament, the Commission and the Council of the European Union. Artificial intelligence is also at the center of a regulatory project in China, the “Measures for Generative Artificial Intelligence Services”. In the run-up to the 2024 American presidential election, Washington is also discussing how these new technologies should be regulated on the Internet.