2026-05-20T22:26:11.835Z / https://www.reuters.com/legal/litigation/trump-sign-order-ai-oversight-security-fears-mount-among-supporters-2026-05-20/
摘要
企业
- 特朗普面临“让美国再次伟大”（MAGA）活动人士施压，要求加强人工智能安全监管
- 科技行业领袖反对强制性人工智能监管，倾向于自愿合作模式
- 关于人工智能监管的辩论可能会影响大模型推出、行业利润与国家安全
华盛顿，5月20日 路透社电——两位知情人士告诉路透社，美国总统唐纳德·特朗普最快将于周四签署一项有关人工智能和网络安全的行政令。目前，其政治基础中的部分群体正施压要求加强对Anthropic公司的Mythos等新型人工智能模型的监管。
消息人士称，该行政令将为人工智能开发商建立一个自愿框架，使其就受监管模型的公开发布与美国政府开展沟通。根据该框架，开发商需在公开发布前90天将模型提交给政府，并向银行等关键基础设施提供商开放预发布访问权限，其中一位消息人士透露。
订阅《每日案卷》新闻简报，将最新法律新闻直接发送至您的收件箱，开启您的晨间资讯。点击此处注册
这种做法可能是特朗普支持者之间的折中方案。
包括特朗普前顾问史蒂夫·班农和右翼政治组织者艾米·克莱默在内的“让美国再次伟大”活动人士，一直在向白宫施压，要求强制人工智能开发商将其最先进的模型提交给政府进行安全测试。
辩论的另一方则是以风险投资家马克·安德森和特朗普前顾问大卫·萨克斯为代表的科技行业支持者，他们反对强制性要求。萨克斯于今年3月辞去特朗普首席人工智能官员一职，目前共同担任总统科技顾问委员会联合主席。特朗普第二任期内的人工智能政策在很大程度上反映了科技行业的立场。
新模型引发辩论
白宫发言人将任何有关人工智能政策细节的讨论称为“猜测”。当被问及总统计划的细节时，美国国家安全局发言人让路透社联系白宫。担任特朗普网络安全政策与战略首席顾问的国家网络主管肖恩·凯恩斯未回应置评请求。
随着包括Mythos和OpenAI的GPT-5.5-Cyber在内的强大新型人工智能系统的发布，特朗普支持者中两大阵营的力量平衡发生了变化。这些公司警告称，新型模型可能会加剧复杂的网络攻击，不过一些网络安全高管表示这些担忧被夸大了。
Mythos的推出引发了特朗普支持者内部就如何应对该技术的斗争。如果特朗普的决定延缓大型语言模型的推出，或促使企业改变模型性能以应对安全担忧，这场辩论的结果可能会对人工智能行业产生重大影响。无论哪种选择都可能损害企业利润。
共和党历来倾向于限制政府权力、反对监管，但越来越多声音响亮的民粹主义共和党支持者呼吁实施人工智能监管。根据上周五他们提交给白宫的一封信，民粹主义派别要求特朗普在“潜在危险”人工智能系统部署前要求获得政府批准。
克莱默表示，倡导新监管规定与她的政治立场“背道而驰”，但人工智能需要采取不同的应对方式。
“你不能指望这些领导人工智能公司的人会把我们的利益放在首位，做正确的事来保护美国人民，”她说。
克莱默曾组织2021年1月6日国会山骚乱前的集会。她在采访中表示，自己并未参与当天数千名特朗普支持者冲击国会大厦的行动。
大型科技公司的支持
科技高管是特朗普最大的政治捐赠者和最显眼的支持者之一。2025年1月特朗普就职时，Meta的马克·扎克伯格、亚马逊的杰夫·贝佐斯、谷歌的桑达尔·皮查伊和OpenAI的山姆·奥尔特曼就坐在前排。
科技行业倡导者告诉路透社，他们希望美国商务部人工智能标准与创新中心在特朗普政府应对先进人工智能模型的过程中发挥主导作用，并补充称企业愿意自愿与该组织的科学家和网络安全专家合作。
另外两位知情人士透露，美国国家安全局与凯恩斯一起参与了政府内部关于如何应对Mythos的讨论。议员们要求凯恩斯与联邦机构合作，建立一个监控“突然出现的前沿人工智能能力跃升”的流程。
“过去几个月给我们敲响了警钟，让我们意识到人工智能可能带来的各类漏洞，”前美国众议员布拉德·卡森说，他目前帮助运营一个超级政治行动委员会网络，其资助者包括Anthropic。
abundance研究所人工智能政策主管尼尔·奇尔森表示，在联邦政府审核期间暂缓推出新人工智能模型，可能会让美国在短期内获得相对于竞争对手的优势，但从长期来看，无法阻止技术落入敌人之手。该非营利组织通常与科技行业立场一致。
“我们需要确保我们正在部署这项技术，并从中获得最大收益，包括强化我们的防御，”奇尔森说。
自愿性的联邦新人工智能模型测试机制已经实施了数年，OpenAI和Anthropic等公司会将其产品提交给人工智能标准与创新中心进行审查。该中心在前总统乔·拜登任期内使用过不同的名称。
商务部于5月宣布，谷歌、xAI和微软已同意提交其人工智能模型接受安全测试，但相关细节随后从其网站上消失。白宫和商务部未回应置评请求，解释为何相关细节被移除。
凯伦·弗莱菲尔德 纽约、考特尼·罗森 华盛顿报道；拉斐尔·萨特、亚历山德拉·阿尔珀 华盛顿、A.J. 文森斯 底特律补充报道；罗德·尼克尔 编辑
我们的准则：汤森路透信托原则
Trump to sign order on AI oversight as security fears mount among supporters
2026-05-20T22:26:11.835Z / https://www.reuters.com/legal/litigation/trump-sign-order-ai-oversight-security-fears-mount-among-supporters-2026-05-20/
Summary
Companies
Trump faces pressure from MAGA activists for stricter AI security oversight
Tech industry leaders resist mandatory AI regulations, favor voluntary collaboration
Debate over AI oversight could affect model rollout, industry profits, and national security
WASHINGTON, May 20 (Reuters) – U.S. President Donald Trump is expected to sign an executive order on AI and cybersecurity as soon as Thursday, two sources familiar with the matter told Reuters, as pressure grows from parts of his political base to increase oversight of new AI models, such as Anthropic’s Mythos.
The order would create a voluntary framework for AI developers to engage with the U.S. government about the public release of covered models, the sources said. Under the framework, the developers would be asked to provide their models to the government 90 days before public release, and also give pre-public access to critical infrastructure providers such as banks, one of the people said.
Jumpstart your morning with the latest legal news delivered straight to your inbox from The Daily Docket newsletter. Sign up here.
Such an approach may represent a middle ground among Trump supporters.
MAGA activists, including former Trump adviser Steve Bannon and right-wing political organizer Amy Kremer, have been pressing the White House to require AI developers to submit their most capable models for government security tests.
On the other side of the debate are tech industry supporters such as venture capitalist Marc Andreessen and former Trump adviser David Sacks, who are resistant to mandatory requirements. Sacks in March stepped down from his role as Trump’s lead AI official and is now co-chairing the president’s tech advisory committee. Trump’s AI policies in his second term have largely reflected the tech industry’s perspective.
NEW MODELS DRIVING DEBATE
A White House spokesperson called any discussion about AI policy details “speculation.” A National Security Agency spokesperson directed Reuters to contact the White House when asked about details of the president’s plan. National Cyber Director Sean Cairncross, who serves as Trump’s principal adviser on cybersecurity policy and strategy, did not respond to requests for comment.
The balance of power between the two groups of Trump’s supporters has shifted, driven by the release of powerful new AI systems, including Mythos and OpenAI’s GPT-5.5-Cyber. The companies warn the new models could supercharge complex cyberattacks, though some cybersecurity executives have said those fears are overblown.
Mythos’ arrival prompted a battle among the president’s supporters to influence how he responds. The outcome of that debate could have a significant impact on the AI industry if the president’s decision slows the rollout of large language models or prompts the companies to change how a model performs to address safety concerns. Either option could hurt profits.
Republicans have traditionally favored limited government and opposed regulations, but support is growing among their more vocal populist supporters to impose AI guardrails. The populist faction is asking Trump to require government approval of “potentially dangerous” AI systems before they’re deployed, according to a letter they sent to the White House last Friday.
Kremer said it is “antithetical” to her political views to advocate for new regulations, but AI requires a different approach.
“You can’t count on these people that are leading these AI companies to put our interests at heart and do what’s right to protect the American people,” she said.
Kremer helped organize a January 6, 2021, rally that preceded the Capitol riot. She said in an interview that she was not among the thousands of Trump supporters who stormed the Capitol that day.
BIG TECH SUPPORT
Tech executives are among the president’s largest political donors and most visible supporters. Meta’s Mark Zuckerberg, Amazon’s Jeff Bezos, Google’s Sundar Pichai and OpenAI’s Sam Altman sat front and center as he was sworn into office in January 2025.
Advocates for the tech industry told Reuters they want to see the U.S. Department of Commerce’s Center for AI Standards and Innovation play a leading role in the Trump administration’s response to advanced AI models, adding that companies are willing to work with the scientists and cybersecurity specialists in that organization voluntarily.
The National Security Agency has been involved in administration-wide discussions about how to respond to Mythos, according to two other people familiar with the matter, along with Cairncross. Lawmakers asked Cairncross to work with federal agencies to set up a process that would monitor “sudden frontier AI capability jumps.”
“The past couple months have served as a massive wake-up call for the kinds of vulnerabilities that AI can create,” said former U.S. Representative Brad Carson, who now helps run a super PAC network whose funders include Anthropic.
Holding back new AI models while the federal government vets them may allow the U.S. to gain a short-term advantage over adversaries but will not keep the technology out of enemy hands in the long term, said Neil Chilson, head of AI Policy at the Abundance Institute, a nonprofit often aligned with the tech industry.
“We need to make sure we’re deploying it and getting the most out of it, including by hardening our defenses,” Chilson said.
Voluntary federal testing of new AI models has been in place for a few years, with companies such as OpenAI and Anthropic
submitting their products for scrutiny by the Center for AI Standards and Innovation, known by a different name under former President Joe Biden.
The Commerce Department announced in May that Google, xAI and Microsoft had agreed to submit their AI models for security testing, though the details later disappeared from its website. The White House and Commerce Department did not respond to requests for comment about why the details disappeared.
Reporting by Karen Freifeld in New York and Courtney Rozen in Washington; Additional reporting by Raphael Satter and Alexandra Alper in Washington and A.J. Vincens in Detroit; Editing by Rod Nickel
Our Standards: The Thomson Reuters Trust Principles.
发表回复