Openai and Google require the government to allow them to train artificial intelligence on their content

Openai and Google pushes the United States government to allow their artificial intelligence models to train on copyrights protected. The two companies set their positions in the proposals published this week, with The controversy opened The application of the protection of fair use to artificial intelligence is “the issue of national security”.

The proposals come in response to the request of the White House, Who asked governmentsIndustry groups, private sector organizations, and others to obtain inputs on the “AI Plan of IQ” for President Donald Trump. The “initiative” is supposed to strengthen America’s position as Amnesty International, while preventing “exhausting requirements” from influencing innovation.

Commenting, the open allegations that allowing artificial intelligence companies to access copyright content would help the United States “avoid losing” its progress in artificial intelligence to China, while Dibsic ascending invitation.

“There is no doubt that the developers of artificial intelligence (the People’s Republic of China) will enjoy unrestricted access to data – including copyright data – that will improve their models,” he writes Openai. “If PRC developers have an unrestricted access to data, and let American companies leave without access to fair use, the artificial intelligence race has ended effectively.”

Google, which is not surprising, agrees. the The company’s response Likewise, it states that copyright, privacy and patent policies “can hinder appropriate access to the data needed to train leading models.” He adds that the policies of just use, along with exceptions to extract text and data, were “decisive” to train artificial intelligence on the data available to the public.

“These exceptions allow the use of copyrights protected and available to the public to train artificial intelligence without greatly affecting the rights holders and avoiding unexpected and unbalanced negotiations often with data holders while developing models or scientific experimentation,” says Google.

Anthropor, AI Chatbot Claud He also made a proposal – But he does not mention anything about copyright. Instead, it requires the US government to develop a system to assess the risks of national security for the artificial intelligence model and enhance export controls on artificial intelligence chips. Like Google and Openai, Anthropor also indicates that the United States enhances energy infrastructure to support the growth of artificial intelligence.