Recently, a short video of American singer Taylor Swift "speaking" Chinese was swiped on the Internet, causing heated discussions.
In the video, Taylor Swift "speaks" fluent Chinese, and his voice is similar to that of his native speaker, and he can even lip sync.
To watch the video, please click on the link below:
AI technology swipes the screen again! What are the risks of celebrities collectively "speaking" foreign languages?
△ The video comes from the netizen @好try
There are many similar videos where foreigners "speak" authentic Chinese, and many Chinese artists "speak" fluent English... Many netizens sighed, "The mouth shape and gestures match, it's so natural, I almost thought it was real." "In addition to the shape of the mouth, there is also the coordination of body movements and speech speed."
In fact, these videos have one thing in common: they are all generated by AI (artificial intelligence).
AI "translates" the real with the fake
I tried an AI portrait video tool, just upload the video, select the language, and generate a new video with automatic translation, timbre adjustment, and lip matching.
△ A video translation function of HeyGen, an AI portrait video tool. Source: Screenshot from HeyGen
Compared with the original film, the new video re-dubbed after AI "translation" not only has a more accurate translation, but also even the voice and emotions are almost realistic, and the mouth shape is very close, making it difficult to distinguish the real from the fake.
In this regard, some netizens said that this technology looks cool, has a wide range of application scenarios, and may become more and more mature in the future, such as translation and dubbing will be more natural and expressive.
But some netizens expressed concern, "Will this be used to imitate the voice to defraud?" "Coupled with the change of face, you won't be able to trust anything on the Internet in the future." "It will be easier and easier to fake videos in the future."
Some netizens said, "I hope that the development of AI technology can be controlled on the right track, and the abuse of technology is a disaster." ”
AI technology is a "double-edged sword"
In recent years, AI technology has been widely used in the secondary creation of short videos, such as AI face swapping and voice swapping. It can be said that AI technology is a "double-edged sword".
On the one hand, AI face swap and voice swap technology has a very strong effect on the improvement of work efficiency. Si Ruo, a professor at Tsinghua University's School of Journalism and Communication, said for example that AI face swap and voice swap technology allows deceased actors and historical figures to be reproduced on the screen through face swap technology, and can also be translated in real time through AI voice swap technology in different countries.
On the other hand, in recent years, some criminals have used AI technology to commit fraud in order to make profits. For example, using AI to swap faces and impersonate acquaintances to defraud, or even extend the "devil's claws" to children, or produce and publish fake news to confuse the public and earn traffic.
△ A blogger recounted a case of traffickers using AI to "swap faces" to abduct children, and the so-called "mother" in the child's video was synthesized by AI. Source: Aunt Jia Child Safety
In addition, there are legal risks associated with AI face swapping and AI voice swapping. Recently, many videos of cross talk actors "speaking" English cross talk and sketch actors "acting" English sketches have spread on the Internet. In this regard, lawyer Yue Daishan, a senior partner of Beijing Yuecheng Law Firm, said that using AI technology to change the voice of others, do "translation" and publish videos may be suspected of infringement, mainly in three aspects:
- First, copyrights, such as cross talk and sketches, are all "works" protected by the Copyright Law of the People's Republic of China. Netizens use AI software to "translate" cross talk, sketches, etc. into other languages, which needs to be authorized by the copyright owner, otherwise there will be infringement problems.
- According to the Civil Code of the People's Republic of China, no organization or individual may infringe upon the portrait rights of others by means of uglification, defacement, or forgery by means of information technology. Without the consent of the portrait rights holder, the portrait of the portrait rights holder must not be produced, used, or disclosed, except as otherwise provided by law. Netizens who use other people's images to make videos and publish them on the website need to obtain the consent of the portrait rights holder, otherwise they are suspected of infringement.
- The third is the right of voice, according to the provisions of the Civil Code of the People's Republic of China, the protection of the voice of natural persons shall be governed by reference to the relevant provisions on the protection of portrait rights. In other words, it is necessary to obtain the consent of the person who has the right to speak before you can use the voice of another person.
Lawyer Yue Daishan emphasized that the abuse of technology may infringe on the rights and interests of others, "AI technology itself is not a problem, the key is how we use it." ”
How to strengthen the regulation of AI technology?
The development of AI technology requires multi-level constraints and norms.
- At the source, it is necessary to further strengthen the protection of citizens' personal information. Experts suggest that the technical and judicial protection of citizens' information, especially biometric and other private information, should be strengthened.
- At the technical level, technologies such as digital watermark counterfeiting need to be further popularized. For example, AI-generated videos need to have an indelible "AI-generated" watermark.
- At the level of the legal system, we will further improve relevant laws and regulations in areas such as artificial intelligence. On August 15 this year, the "Interim Measures for the Management of Generative AI Services" came into effect. The Measures stipulate that the provision and use of generative AI services shall respect the legitimate rights and interests of others, shall not endanger the physical and mental health of others, and shall not infringe upon the rights and interests of others' portrait, reputation, honor, privacy and personal information.
Zuo Xiaodong, a professor at the School of Public Affairs and the School of Cyberspace Security at the University of Science and Technology of China, said that in general, generative AI is still in the early stage of development, the risks are not fully exposed, and the application prospects are difficult to estimate, so we should adopt an inclusive and prudent attitude, adhere to the combination of development and security, promote innovation and governance according to law, and take effective measures to encourage the innovative development of generative AI.
It is understood that at present, the artificial intelligence industry in mainland China is booming, with the scale of the core industry reaching 500 billion yuan and the number of enterprises exceeding 4,300. With the introduction of various measures and norms, we look forward to the development of artificial intelligence in the future to be more healthy and standardized, and help people's work and life in many aspects.
▌Source of this article: CCTV News WeChat public account (ID: cctvnewscenter)