China reveals draft deepfake regulations that restrict use • The Register

The Chinese government has unveiled a draft law clamping down on deepfakes – the practice of using AI to adapt existing digital content into realistic simulations of humans.

The draft emerged last Friday from the Cyberspace Administration of China and frames the need for regulation in the context of the government’s desire to ensure the internet is a tool for good and not the wretched hive of scum and villainy it has often become.

The explanatory memorandum for the policy suggests criminals and fraudsters will be attracted to using digitally created voice, video, chatbots, or manipulation of faces or gestures. The draft therefore rules out the use of such fakes for any application that could disrupt social order, infringe individuals’ rights, deliver fake news, or depict sexual activity. It also proposes requiring a grant of permission for use of what China calls “deep synthesis” before it can be employed for legitimate uses.

Just what those legitimate uses might be is not discussed, but the draft does outline extensive regulations on how digital assets must be safeguarded to prevent user privacy.

If deep synthesis is used, the draft proposes a requirement for it to be marked as a digital creation to remove any doubt about authenticity and provenance.

The draft also outlines requirements for service providers to implement excellent security practices and always act in the national interest.

The Middle Kingdom’s big tech companies got the same message on the weekend, during a symposium on promoting the healthy and sustainable development of internet companies. The heads of 27 companies attended the event, at which Chinese regulators explained their desire for internet platforms to be both innovative problem-solvers and steadfast defenders of Chinese values.

Zhuang Rongwen, deputy director of China’s Central Propaganda Department, called on big tech to ensure it continues to strengthen Chinese society with brilliant online services, while at the same time stepping up vigilance to ensure the Chinese internet is free of the many types of content that Beijing believes are bad for society.

Leaders from the People’s Daily Online, JD.com, Kuaishou, Xiaomi, and Meituan chimed in with their views on how big tech companies can ensure China’s internet conforms to the Party line, but sadly their remarks were omitted from the Administration’s account of the event. ®