Regulators in China have unveiled new draft rules to manage how companies develop artificial intelligence tools like ChatGPT.
The draft rules by the Cyberspace Administration of China seek to manage fast-developing Generative AI tools after tech giants like Alibaba and Baidu announced their plans to roll out their own version of the AI chatbot.
Generative AI tools are trained to produce unique human-like content such as images and texts by analysing large quantities of data.
After the American AI lab OpenAI launched ChatGPT last year in November, it gained widespread recognition for its ability to respond to user queries with human-like language.
It has demonstrated a range of abilities, from summarising complex research studies in plain language to answering logical questions, and has also cracked business school and medical exams crucial for students to pass, prompting several companies to unveil their own versions of generative AI chatbots.
Alibaba revealed on Tuesday that it too is working on a ChatGPT rival Tongyi Qianwen that it planned to integrate across its services.
Google and the Chinese company Baidu have also revealed they are working on AI chatbots.
Following these moves in the tech industry, China’s cyberspace regulator has released draft measures for managing generative AI in the country.
The draft measures lay the ground rules for companies developing generative AI services to follow such as the type of content their products can generate.
Companies are expected to ensure that the data used to train the AI models do not discriminate against people on the basis of race, gender, and ethnicity, and they would also be responsible for the legitimacy of the data used to train the algorithm, the new rules noted.
The AI Chatbots developed by Chinese companies should not generate false information, the regulator noted.
Several AI experts have warned that chatbots like ChatGPT can sometimes provide plausible-sounding but incorrect responses with glaring mistakes to some queries.
They say it may confidently provide incorrect answers that may mislead people, adding that its widespread use may also result in copyright infringement.
Recently several Japanese universities, including Tokyo’s Sophia University, restricted students’ use of the OpenAI chatbot over concerns about information leaks from the use of the AI tool.
Italy has also banned the use of ChatGPT with the AI service set to be investigated over privacy concerns.
Germany also noted last week that it is considering a ban on ChatGPT over privacy concerns.
