China says chatbots must toe the party line

Five months after ChatGPT sparked an artificial intelligence investment frenzy, Beijing is set to curb Chinese chatbots, evidence of the government’s determination to maintain tight regulatory control over the technology that could define an era.

China’s Cyberspace Administration this month unveiled draft rules for so-called generative artificial intelligence — software systems, like the one behind ChatGPT, that can formulate text and images in response to questions and prompts. ‘an user.

According to the regulations, companies must abide by the Chinese Communist Party’s strict censorship rules, just as websites and apps must avoid posting material that tarnishes Chinese leaders or resurrects banned history. The content of AI systems should reflect “core socialist values” and avoid information that undermines “state power” or national unity.

Companies will also need to ensure that their chatbots create truthful, intellectual property-friendly words and images, and will need to register their algorithms, the software brains behind the chatbots, with regulators.

The rules aren’t final and regulators may continue to tweak them, but experts said engineers building artificial intelligence services in China are already looking to incorporate the decrees into their products.

Governments around the world have been impressed by the power of chatbots with AI-generated results ranging from alarming to benign. Artificial intelligence has been used to pass university exams and create a fake photo of Pope Francis in a puffy coat.

ChatGPT, developed by US-based OpenAI, which is backed by some $13 billion from Microsoft, has spurred Silicon Valley to apply the underlying technology to new areas like video games and advertising. Venture capital firm Sequoia Capital estimates that AI companies could eventually produce “trillions of dollars” in economic value.

In China, investors and entrepreneurs are racing to catch up. Shares of Chinese artificial intelligence companies have soared. Loud announcements have been made by some of China’s biggest tech companies, most recently including e-commerce giant Alibaba; SenseTime, which makes facial recognition software; and the Baidu search engine. At least two startups developing Chinese alternatives to OpenAI technology have raised millions of dollars.

ChatGPT is not available in China. But faced with a growing number of local alternatives, China was quick to unveil its red lines for artificial intelligence, ahead of other countries still mulling over how to regulate chatbots.

The rules showcase China’s approach to regulation “move fast and break things,” said Kendra Schaefer, head of technology policy at Trivium China, a Beijing-based consultancy.

“Because you don’t have a two-party system where both sides argue, they can just say, ‘OK, we know we have to do this, and we’ll revise it later,’” she added.

Chatbots are trained across large swathes of the internet, and developers are grappling with inaccuracies and surprises in what they sometimes spit out. At first glance, Chinese rules require a level of technical control over chatbots that Chinese tech companies have not achieved. Even companies like Microsoft are still tweaking their chatbots to weed out harmful responses. China has a much higher bar, which is why some chatbots have already been shut down and others are only available to a limited number of users.

Experts are divided on the difficulty of training AI systems to always be factual. Some doubt the companies can account for the range of Chinese censorship rules, which are often sweeping, constantly changing and even require censorship of specific words and dates like June 4, 1989, the day of the Tiananmen Square massacre. Others believe that over time and with enough work, machines can be aligned with truth and specific, even political, value systems.

Analysts expect the rules to undergo changes after consultation with Chinese tech companies. Regulators could loosen their enforcement so that the rules don’t completely undermine the development of the technology.

China has a long history of internet censorship. Throughout the 2000s, the country built the world’s strongest net of information on the web. This has scared away non-compliant Western companies like Google and Facebook. He hired millions of workers to monitor internet activity.

Meanwhile, Chinese tech companies, which had to abide by the rules, thrived, defying Western critics who predicted political control would undermine growth and innovation. As technologies such as facial recognition and cellphones emerged, corporations helped the state harness them to create a surveillance state.

The current wave of AI poses new risks for the Communist Party, said Matt Sheehan, Chinese AI expert and fellow at the Carnegie Endowment for International Peace.

The unpredictability of chatbots, which will make absurd or false statements – what AI researchers call hallucinations – runs counter to the party’s obsession with managing what is said online, Mr Sheehan said.

“Generative artificial intelligence has put two of the party’s main goals in tension: control of information and leadership in artificial intelligence,” he added.

China’s new regulations aren’t just about politics, experts said. For example, they aim to protect the privacy and intellectual property of individuals and creators of the data on which AI models are trained, a matter of global concern.

In February, Getty Images, the image database company, sued artificial intelligence startup Stable Diffusion for training its image generation system on 12 million watermarked photos, which it claims Getty, diluted the value of its images.

China is making a broader effort to answer legal questions about AI companies’ use of underlying data and content. In March, as part of a major institutional overhaul, Beijing created the National Data Office, an effort to better define what it means to own, buy and sell data. The state body would also help companies create the datasets needed to train such models.

“They are now deciding what kind of proprietary data is and who has the right to use and control it,” said Schaefer, who has written extensively on Chinese AI regulations and called the initiative a “transformer”.

Still, the new Chinese railings might be inappropriate. The country faces intensifying competition and semiconductor sanctions that threaten to undermine its competitiveness in technologies, including artificial intelligence.

Hopes for Chinese AI were high in early February when Xu Liang, an engineer and artificial intelligence entrepreneur, released one of the first Chinese responses to ChatGPT as a mobile app. The app, ChatYuan, garnered more than 10,000 downloads in the first hour, Xu said.

Media reports of stark differences between the party line and ChatYuan’s responses soon surfaced. The responses offered a grim diagnosis of the Chinese economy and described Russia’s war in Ukraine as a “war of aggression”, at odds with the party’s more pro-Russian stance. A few days later, authorities shut down the app.

Mr. Xu said he was adding metrics to create a more “patriotic” bot. They include filtering out sensitive keywords and hiring more manual reviewers who can help flag problematic responses. It even trains a separate model capable of detecting “incorrect viewpoints”, which it will filter out.

Still, it’s unclear when Mr. Xu’s bot will ever satisfy the authorities. The app was originally supposed to resume on February 13, according to screenshots, but on Friday it was still down.

“Service will resume once troubleshooting is complete,” it read.