保守派 AI 安全組織 Humans First 提出「America First AI Policy」綱領，倡議收緊出口管制及暫停 H-1B 簽證
Conservative AI safety group proposes export controls, visa restrictions
保守派主導的組織 Humans First 提出「America First AI Policy」綱領，倡議收緊對中國及其他對手的出口管制，並暫停 H-1B 簽證。綱領亦要求大型 AI 公司接受獨立測試、公開通報 AI 事故，禁止開發遞歸式自我改進等先進 AI 做法，並主張追究前沿 AI 開發者魯莽行為的責任、為兒童使用 AI 設嚴格規則，以及禁止聯邦政府使用 Flock 等攝影機。
Humans First, a conservative-led group advocating for stronger AI guardrails, is out with an “America First AI Policy” platform on Friday, according to a memo shared first with Semafor.
Key parts of the framework include tightening export controls against China and other adversaries; mandating independent testing for large AI companies; prohibiting the development of advanced AI practices like recursive self-improvement; banning federal use of cameras like Flock; requiring public reporting of AI incidents; and ensuring that frontier AI developers can be held liable for reckless conduct. The group also wants strict rules for AI used by children and calls for a pause on H-1B visas.
“An AI policy that truly puts America First is centered around the American people — not globalist corporations and not unaccountable bureaucrats,” said Russ Walker, the group’s executive director and an adviser to President Donald Trump’s 2020 campaign. “It is time for Congress to act and it is time for conservatives in particular to lead on this critical issue.”
The group is chaired by Amy Kremer, a longtime Tea Party activist who argues Republicans are missing an opportunity to lead on AI policy that doesn’t cater to big companies. She’s been leading a bus tour across the country dubbed the “AI Data Center Revolt.”
The framework’s priorities go beyond what some Republicans who are interested in regulating AI on Capitol Hill have proposed, and significantly further than the Trump administration’s current light touch on AI.
Still, this vocal group may have more sway going forward if Republicans perform poorly in the midterms and AI anxiety is a factor.
來源：Semafor · semafor.com