danger to democracy
Europe and US want to restrict China’s “ideological” AI
4/13/2023 9:54 p.m
China wants to support the development of AI. However, this must correspond to basic socialist values. The announcement causes concern in Europe and the United States. The call for regulation or bans is getting louder.
China’s planned mandates for an ideological orientation of its AI systems have fueled calls in the West for regulation of the rapidly evolving industry. Austrian State Secretary Florian Tursky, who is responsible for digitization, said that if the plans were to be implemented, “it would be necessary to exclude AI systems created in China from the EU market and to ban them in Europe”. Majority leader in the US Senate, Chuck Schumer, also referred to China, speaking of the need to act quickly given the impact of artificial intelligence (AI) “on our society and the global economy”. He is preparing a non-partisan initiative.
China’s Internet regulator CAC announced on Tuesday that it would support the development of generative AI – the variant of the technology behind ChatGPT, for example. However, the corresponding content would have to correspond to the basic socialist values. Providers should therefore be responsible for the data and should expect penalties if they do not comply with the rules.
“Ideological Footprints of the Communist Party”
Trusky called the Chinese announcement a threat to democracy. “AI must not follow any state-imposed ideology. This would result in Chinese AI systems coming to Europe that have the ideological footprints of the Chinese Communist Party.” Trusky has therefore asked the EU Commission to complete the set of rules for the Artificial Intelligence Act as quickly as possible. It is intended to ensure that the potential of AI can be safely exploited. One cannot wait until early 2025 for EU regulation, Trusky explained. Two years ago, the EU Commission presented a first draft law with the declared aim of protecting citizens from the risks of AI. Because of the rapid development, however, hopes of a compromise at EU level were dashed.
According to his office, US Democrat Schumer has been discussing an “overriding framework” for regulation for months. Leading AI experts were also involved. The aim is to provide specifications that can be adapted to further developments. There was initially no comment from Republicans. They hold the majority in the House of Representatives and have traditionally been skeptical about state regulations. On Tuesday, President Joe Biden’s government took the first steps towards possible guidelines: An advisory committee in the Department of Commerce asked for public statements on an accountability mechanism for AI.
Generative artificial intelligence has made waves in politics and business through ChatGPT of the Microsoft holding OpenAI. The program can, for example, answer inquiries or create complete texts based on a few keywords. Large, pre-fed amounts of text serve as the basis. There are concerns about incorrect answers and data protection. ChatGPT is blocked in Italy, France’s data protection officer CNIL announced an examination. The Spanish sister authority AEPD has asked the EU authorities to take appropriate steps. In Germany, the Federal Digital Ministry rejects a ban.