Google Call for copyright and exporting the rules on the AI ​​policy proposal

Google, below in the Opequai heels, issuing a policy proposal for Trump administrative calls for the action of the “AI.” Tech giants push copyright limitations in AI training, as well as “balanced” Control the “protecting national security when enables global business operations and operations.” “The US should be a continuing international economic policy of the American economy and supports an international innovation,” Google Write in the document. “For a long time, AI policy has paid attention, ignoring customized charges with regulations, and national leadership – dynamic beginning according to new administration.” One of Google’s controversial recommendations are more regarding the use of IP protected material. Google is emphasized that “mining and text-and-data” existing “is” critical “for ai innovation related AI related to AI related to ai. Like the opening, the company wants to accommodate the rights and rivals train in public data – including copyright data – generally unlimited. “The exception allows copyright use, in general for the AI ​​training without negotiation,” and so uncomfortable, and unbalanced and imbalance and scientific experiments. “Google, who have been trained with several models in copyrighted data, the US owners are performed by the administration of an offer, which may break the rival load” can eliminate the Unabled load in the US Cloud Service. “The statement of the Google’s strangers like Microsoft, which in January said” belief “can” fully “with the rules. It is important, export rules, which seemingly limit the availability of AI Ai progresses, carved exemptions for a believable business to find a trusted cluster. On the other place, Google calls “long-term” on investment in the new Federal R & D, pressure to reduce the throw and eliminate award. The company says that the government has to release datasents that can be useful for the commercial AI training, and provides funds for “early market r & d” when it ensures a “multiple model”. Pointing to the chaos-based environment created by the United States Patchwork Act, Google get the government to pass the Federal rules in AI, including complete safety and security frames and security. Only two months to be 2025, the number of AI bills pending in the US has grown up to 781, according to an online tracking tool. Google benefits the US government fighting for what we are about to be an obligation around the AI ​​system, like the obligation of use responsibility. In many cases, Google states, model developers “No visibility or control” on how the model used and so it cannot misuse. History, Google has won the law like California beats SB 1047, which must be clearly applied that the AI ​​developer should be held to harm affected. “Even in the case where developers provide a model directly to representatives, devloyers can often understand the hille usage of the risk, and the rescue of market delivery and logs,” Google Write. Google in the proposal also mentioned the terms as they are considered “very big,” the Government of the Commerce required to refer to the product to refer to or the Jailbreak Model. “Many countries and countries that have passed the law of ai to reveal the AI ​​system if the corner of the formation of the formation of instructions, restrictions, and risks associated with the model.

Source link