Science and technology groups urge the Trump White House to retain the “key rules” of AI | Real Time Headlines

According to a letter from the US Consumer Federation and Mozilla, the White House should retain the “key rules” of artificial intelligence testing and transparency.

This letter is viewed by CNBC, and then President Donald Trump Decide to withdraw the decision of the former president Biden 2023 AI’s administrative order New security assessment, fairness and civil rights guidance, and research on the impact of AI on the labor market. It is a speech on several officials including the White House AI and Crypto Tsar, as well as the new national security consultant Mike Waltz.

Biden’s command stipulates that developers of large -scale AI systems, especially developers who may be risky for national security, public health or economic risks. Before publishing this technology to the public, please share security tests with the US government. result.

The signing of this letter, including the digital democratic center and the National Consumer Legal Center, pointed out that President Trump’s administrative order will revise the rules and “require the federal government to test and disclose the AI ​​system before using consumers.” They It is said that these systems are used to help senior affairs department give priority to nursing and review retirement benefits.

“Without the test of the fences on the test and transparency on the AI ​​system, the foundation of the guardrail is so basic, so that any engineer should feel ashamed to release the product without them-the elderly, veterans and veterans and Consumers will suffer improper benefits, and their health is endangered, “they are endangered,” wrote. “

They said that the standards set in the previous rules were “not very high”, “it is our elderly, veterans, and daily consumers.”

After Biden’s command, many folk social leaders praised this 111 -page document as a step in the correct direction in the correct direction. Not far away Identification and resolution from the real world hazards from the AI ​​model. But many technological leaders worry that these rules will hinder innovation. Nevertheless, it is still widely considered Effective compromiseEssence

Due to potential harmful chain reactions, AI has been controversial for a long time, especially for fragile and minority groups. Police use AI to cause a lot Error arrestThe survey shows that the car insurance algorithm will Oppose marginal communityResearch found Major racial differences In mortgage underwriting

The organization participating in the writing of Thursday said that the security rules of the pre -administrative order are suitable for the establishment of a large enterprise of the AI ​​system, which affects “the most vulnerable person.” They said that the use of taxpayers in the unliked AI system may lead to “further waste, fraud and abuse.”

This letter says: “The problem we emphasize here is not about ‘ideological prejudice’ or ‘engineering social agenda’, as President Trump’s latest administrative orders on AI.” “On the contrary, the problem here is about the problem. The basic principles of safety engineering are critical to the technology responsible for using all other technologies that affect millions of people, from how we test the aircraft to how we protect the software. “

watch: Deep flavor and distillation