China has added a new one to its breakthroughs in the field of artificial intelligence. China, which previously launched a big competition in language processing with the DeepSeek model, is now making an important move in the field of video production. TikTok’s roof company ByteDance announced a new artificial intelligence model called OmniHuman-1. This model can produce videos that can show people talking, dancing, singing or playing a musical instrument with only an image or text input.

OmniHuman-1 will rival OpenAI’s Sora

OmniHuman-1 was introduced as a multi-mode artificial intelligence model that can mimic human movements, speech and gestures very close to reality in video production. According to the company, this model can produce natural movements and realistic expressions, whether it is a portrait photo or a full-body image. This technology is seen as a direct competitor to the Sora model introduced by OpenAI last December.

Key features of OmniHuman-1 are:

• Realistic human gestures and facial expressions
• Multiple input support (images, audio and video tracks)
• High quality and realistic video production
• Potential to be used for non-human animations…

Let’s add this: Such advanced artificial intelligence technologies also raise ethical and security concerns. Some experts point to the potential for misuse of such models, saying: ‘Such artificial intelligence models can facilitate the production of deepfake content, which can lead to information pollution. Therefore, the development and use of these technologies should be strictly controlled.’

In this context, some academics and ethicists argue that models like OmniHuman-1 should be banned or strictly regulated due to their potential risks.

ByteDance’s breakthrough in this area could further increase the role of AI in video content production in the future. How do you think China’s aggressive AI moves will change the industry? Share your views in the comments!