<p><img src="https://imageproxy.pbkrs.com/https://invezz-wp-media.lon1.digitaloceanspaces.com/2026/04/image-1776856218.png?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" alt="" original-src="https://imageproxy.pbkrs.com/https://invezz-wp-media.lon1.digitaloceanspaces.com/2026/04/image-1776856218.png"/></p>
<p>主要经济体的金融监管机构正在加强对 Anthropic 开发的一种强大新人工智能模型的审查，因为人们越来越担心其先进能力可能会暴露关键金融和数字系统的脆弱性。</p>
<p>日本、欧洲、印度等地区的当局正在协调应对，因为这一名为 Mythos 的模型引发了关于在日益依赖人工智能的世界中网络安全风险的新问题。</p>
<h2>日本召开顶级银行和监管机构会议</h2>
<p>日本金融服务局周三表示，将于周五与包括三菱 UFJ 金融集团、住友三井金融集团和瑞穗金融集团在内的国家最大银行集团举行高级别会议。</p>
<p>此次会议还将包括日本银行和东京证券交易所，反映出金融系统普遍存在的担忧。</p>
<p>片山聪子表示，会议的目的是汇集关键利益相关者，以评估不断变化的形势。</p>
<p>&#34;我们将汇集承担最大责任的核心成员，分享对当前形势的评估并交换意见，包括国际金融界各个部分提出的问题，&#34;她说。</p>
<h2>全球中央银行保持警惕</h2>
<p>其他地区的监管机构也采取了类似措施。</p>
<p>澳大利亚储备银行周三表示，正在密切关注事态发展，并&#34;与同行监管机构、政府和受监管实体进行接触。&#34;</p>
<p>与此同时，新西兰储备银行周三将风险描述为 “正在发展中”，并确认正在与国内和澳大利亚的同行进行协调。</p>
<p>印度也加入了全球努力。</p>
<p>据路透社报道，印度储备银行正在与国际监管机构、贷款机构和政府官员进行讨论，以评估潜在威胁。</p>
<p>初步评估表明，该模型可能加速软件漏洞的发现和利用，增加网络安全隐患。</p>
<p>这一问题已上升到全球政策制定的最高层面。</p>
<p>加拿大财政部长弗朗索瓦 - 菲利普·香槟告诉 BBC，该模型在最近的国际货币基金组织华盛顿会议上进行了讨论。</p>
<p>&#34;当然，这个问题足够严重，值得所有财政部长的关注，&#34;他说，并将该技术形容为 “未知的未知”。</p>
<p>同样，英格兰银行行长安德鲁·贝利警告说，监管机构正在密切评估其影响。</p>
<p>“我们现在必须非常仔细地审视这一最新的人工智能发展可能对网络犯罪风险意味着什么，” 他告诉 BBC。</p>
<h2>网络安全的双刃剑</h2>
<p>尽管 Mythos 是为防御性网络安全应用而设计的，但其能力引发了警觉。</p>
<p>Anthropic 表示，早期测试显示在主要操作系统和网络浏览器中发现了 “数千个” 重大漏洞。</p>
<p>周二，德国联邦银行行长约阿希姆·纳格尔将该模型形容为 “双刃剑”。</p>
<p>&#34;它不仅可以用于改善数字安全系统，还可以利用其脆弱性进行恶意用途。&#34;</p>
<p>专家表示，令人担忧的是这种人工智能系统识别和利用弱点的速度和规模，远远超过传统的安全响应。</p>
<h2>为什么 Anthropic 限制对该模型的访问？</h2>
<p>Anthropic 通过一个名为 Project Glasswing 的受控计划限制了对该模型的访问。</p>
<p>包括亚马逊、微软、英伟达和苹果在内的部分组织已获得早期访问权限，以及负责维护关键数字基础设施的数十个实体。</p>
<p>在一段发布视频中，达里奥·阿莫代伊表示，公司已主动提出与美国当局合作，以 “帮助防御这些模型的风险”。</p>
<p>包括所谓的 “红队” 在内的独立研究人员的测试突显了该模型的先进能力。</p>
<p>根据 Anthropic 的说法，Mythos 可以自主识别遗留系统中的关键漏洞，甚至建议利用这些漏洞的方法。</p>
<p>&#34;Mythos Preview 已经发现了数千个高严重性漏洞，包括每个主要操作系统和网络浏览器中的一些漏洞，&#34;该公司表示。</p>
<p>&#34;考虑到人工智能进展的速度，这种能力的传播不会太久，可能超出那些致力于安全部署的行为者。&#34;</p>
<h2>遗留系统的风险最高</h2>
<p>潜在威胁在依赖复杂和老旧基础设施的行业中尤为严重。</p>
<p>贝恩公司的一份报告指出，能源、制造和运输等行业可能因难以修补的过时系统而面临更高的风险。</p>
<p>银行也面临脆弱性，因为它们依赖于相互连接的系统，其中一些系统可以追溯到几十年前。</p>
<p>然而，该报告强调，凭借强大的网络安全实践，这些风险是可控的。</p>
<p>&#34;威胁是严重的，但并非不可克服，强大的网络安全基础是你最好的防御，&#34;报告指出，并补充说，现有的保护措施如网络分段、访问控制和异常检测可以提供显著保护。</p>
<h2>关注准备而非恐慌</h2>
<p>专家认为，与其急于开发全新的人工智能特定防御，组织应优先加强现有的安全框架。</p>
<p>建议包括建立专门的人工智能威胁响应团队、改善基础网络安全措施以及解决操作技术环境中的漏洞。</p>
<p>随着对强大人工智能系统的访问扩大，监管机构和公司面临着平衡创新收益与防范意外后果的挑战。</p>
<p>对 Mythos 的迅速全球反应表明，政策制定者对风险有着敏锐的意识。</p>
<p>这些早期的努力是否足以减轻下一代人工智能带来的风险仍然是一个悬而未决的问题。</p>
<p>文章《为什么 Anthropic 的 Mythos AI 让监管机构和中央银行感到紧张》最早出现在 Invezz 上。</p>

3042

<p>全球金融监管机构正加强对 Anthropic 的 AI 模型 Mythos 的审查，原因是担心它可能在金融系统中暴露的网络安全漏洞。日本金融服务局正在召集主要银行集团评估风险，而澳大利亚、新西兰和印度的中央银行也在监测相关进展。该模型旨在用于网络安全，但因其可能利用软件漏洞而引发警报。随着对其对网络犯罪和传统系统影响的担忧加剧，Anthropic 已限制对该模型的访问，仅允许部分组织进行测试</p>

<p><img src="https://imageproxy.pbkrs.com/https://invezz-wp-media.lon1.digitaloceanspaces.com/2026/04/image-1776856218.png" alt="" original-src="https://invezz-wp-media.lon1.digitaloceanspaces.com/2026/04/image-1776856218.png"/></p>
<p>Financial regulators across major economies are stepping up scrutiny of a powerful new artificial intelligence model developed by Anthropic, amid growing concerns that its advanced capabilities could expose vulnerabilities in critical financial and digital systems.</p>
<div class="lb-trans"><p>主要经济体的金融监管机构正在加强对 Anthropic 开发的一种强大新人工智能模型的审查，因为人们越来越担心其先进能力可能会暴露关键金融和数字系统的脆弱性。</p>
</div><p>Authorities in Japan, Europe, India, and other regions are coordinating responses as the model, known as Mythos, raises fresh questions about cybersecurity risks in an increasingly AI-driven world.</p>
<div class="lb-trans"><p>日本、欧洲、印度等地区的当局正在协调应对，因为这一名为 Mythos 的模型引发了关于在日益依赖人工智能的世界中网络安全风险的新问题。</p>
</div><h2>Japan convenes top lenders and regulators</h2>
<div class="lb-trans"><h2>日本召开顶级银行和监管机构会议</h2>
</div><p>In Japan, the Financial Services Agency on Wednesday said it will hold a high-level meeting on Friday with the country’s largest banking groups, including Mitsubishi UFJ Financial Group, Sumitomo Mitsui Financial Group, and Mizuho Financial Group.</p>
<div class="lb-trans"><p>日本金融服务局周三表示，将于周五与包括三菱 UFJ 金融集团、住友三井金融集团和瑞穗金融集团在内的国家最大银行集团举行高级别会议。</p>
</div><p>The gathering will also include the Bank of Japan and the Tokyo Stock Exchange, reflecting the breadth of concern across the financial system.</p>
<div class="lb-trans"><p>此次会议还将包括日本银行和东京证券交易所，反映出金融系统普遍存在的担忧。</p>
</div><p>Satsuki Katayama said the aim was to bring together key stakeholders to assess the evolving situation.</p>
<div class="lb-trans"><p>片山聪子表示，会议的目的是汇集关键利益相关者，以评估不断变化的形势。</p>
</div><p>&#34;We will bring together the core members who bear the greatest responsibility to share assessments of the current situation and exchange views, including issues that have been flagged in various parts of the international financial community,&#34; she said.</p>
<div class="lb-trans"><p>&#34;我们将汇集承担最大责任的核心成员，分享对当前形势的评估并交换意见，包括国际金融界各个部分提出的问题，&#34;她说。</p>
</div><h2>Central banks globally on alert</h2>
<div class="lb-trans"><h2>全球中央银行保持警惕</h2>
</div><p>Regulators in other regions are taking similar steps.</p>
<div class="lb-trans"><p>其他地区的监管机构也采取了类似措施。</p>
</div><p>The Reserve Bank of Australia on Wednesday said it was closely monitoring developments and &#34;engaging with peer regulators, government and regulated entities.&#34;</p>
<div class="lb-trans"><p>澳大利亚储备银行周三表示，正在密切关注事态发展，并&#34;与同行监管机构、政府和受监管实体进行接触。&#34;</p>
</div><p>Meanwhile, the Reserve Bank of New Zealand on Wednesday described the risks as “developing” and confirmed it was coordinating with domestic and Australian counterparts.</p>
<div class="lb-trans"><p>与此同时，新西兰储备银行周三将风险描述为 “正在发展中”，并确认正在与国内和澳大利亚的同行进行协调。</p>
</div><p>India has also joined the global effort.</p>
<div class="lb-trans"><p>印度也加入了全球努力。</p>
</div><p>According to a Reuters report, the Reserve Bank of India is in discussions with international regulators, lenders, and government officials to evaluate potential threats.</p>
<div class="lb-trans"><p>据路透社报道，印度储备银行正在与国际监管机构、贷款机构和政府官员进行讨论，以评估潜在威胁。</p>
</div><p>Preliminary assessments suggest the model could accelerate the discovery and exploitation of software vulnerabilities, raising cybersecurity concerns.</p>
<div class="lb-trans"><p>初步评估表明，该模型可能加速软件漏洞的发现和利用，增加网络安全隐患。</p>
</div><p>The issue has reached the highest levels of global policymaking.</p>
<div class="lb-trans"><p>这一问题已上升到全球政策制定的最高层面。</p>
</div><p>Canadian finance minister François-Philippe Champagne told the BBC the model had been discussed at a recent International Monetary Fund meeting in Washington.</p>
<div class="lb-trans"><p>加拿大财政部长弗朗索瓦 - 菲利普·香槟告诉 BBC，该模型在最近的国际货币基金组织华盛顿会议上进行了讨论。</p>
</div><p>&#34;Certainly it is serious enough to warrant the attention of all the finance ministers,&#34; he said, describing the technology as an &#34;unknown unknown&#34;.</p>
<div class="lb-trans"><p>&#34;当然，这个问题足够严重，值得所有财政部长的关注，&#34;他说，并将该技术形容为 “未知的未知”。</p>
</div><p>Similarly, Bank of England chief Andrew Bailey warned that authorities are closely evaluating the implications.</p>
<div class="lb-trans"><p>同样，英格兰银行行长安德鲁·贝利警告说，监管机构正在密切评估其影响。</p>
</div><p>“We are having to look very carefully now at what this latest AI development could mean for the risk of cyber crime,” he told the BBC.</p>
<div class="lb-trans"><p>“我们现在必须非常仔细地审视这一最新的人工智能发展可能对网络犯罪风险意味着什么，” 他告诉 BBC。</p>
</div><h2>A double-edged sword for cybersecurity</h2>
<div class="lb-trans"><h2>网络安全的双刃剑</h2>
</div><p>Although Mythos was designed for defensive cybersecurity applications, its capabilities have sparked alarm.</p>
<div class="lb-trans"><p>尽管 Mythos 是为防御性网络安全应用而设计的，但其能力引发了警觉。</p>
</div><p>Anthropic said early testing revealed “thousands” of significant vulnerabilities across major operating systems and web browsers.</p>
<div class="lb-trans"><p>Anthropic 表示，早期测试显示在主要操作系统和网络浏览器中发现了 “数千个” 重大漏洞。</p>
</div><p>On Tuesday, Bundesbank President Joachim Nagel described the model as a “double-edged sword.”</p>
<div class="lb-trans"><p>周二，德国联邦银行行长约阿希姆·纳格尔将该模型形容为 “双刃剑”。</p>
</div><p>&#34;It could be used not only to improve digital security systems, but also to leverage their vulnerabilities for malicious purposes.&#34;</p>
<div class="lb-trans"><p>&#34;它不仅可以用于改善数字安全系统，还可以利用其脆弱性进行恶意用途。&#34;</p>
</div><p>Experts say the concern lies in the speed and scale at which such an AI system can identify and exploit weaknesses—far outpacing traditional security responses.</p>
<div class="lb-trans"><p>专家表示，令人担忧的是这种人工智能系统识别和利用弱点的速度和规模，远远超过传统的安全响应。</p>
</div><h2>Why has Anthropic restricted access to the model?</h2>
<div class="lb-trans"><h2>为什么 Anthropic 限制对该模型的访问？</h2>
</div><p>Anthropic has restricted access to the model through a controlled initiative called Project Glasswing.</p>
<div class="lb-trans"><p>Anthropic 通过一个名为 Project Glasswing 的受控计划限制了对该模型的访问。</p>
</div><p>Select organisations, including Amazon, Microsoft, Nvidia, and Apple, have been granted early access, alongside dozens of entities responsible for maintaining critical digital infrastructure.</p>
<div class="lb-trans"><p>包括亚马逊、微软、英伟达和苹果在内的部分组织已获得早期访问权限，以及负责维护关键数字基础设施的数十个实体。</p>
</div><p>In a launch video, Dario Amodei said the company had offered to collaborate with US authorities to “help defend against the risk of these models.”</p>
<div class="lb-trans"><p>在一段发布视频中，达里奥·阿莫代伊表示，公司已主动提出与美国当局合作，以 “帮助防御这些模型的风险”。</p>
</div><p>Testing by independent researchers, including so-called “red teams,” has highlighted the model’s advanced capabilities.</p>
<div class="lb-trans"><p>包括所谓的 “红队” 在内的独立研究人员的测试突显了该模型的先进能力。</p>
</div><p>According to Anthropic, Mythos can autonomously identify critical bugs in legacy systems and even suggest methods to exploit them.</p>
<div class="lb-trans"><p>根据 Anthropic 的说法，Mythos 可以自主识别遗留系统中的关键漏洞，甚至建议利用这些漏洞的方法。</p>
</div><p>&#34;Mythos Preview has already found thousands of high-severity vulnerabilities, including some in every major operating system and web browser,&#34; the company said.</p>
<div class="lb-trans"><p>&#34;Mythos Preview 已经发现了数千个高严重性漏洞，包括每个主要操作系统和网络浏览器中的一些漏洞，&#34;该公司表示。</p>
</div><p>&#34;Given the rate of AI progress, it will not be long before such capabilities proliferate, potentially beyond actors who are committed to deploying them safely.&#34;</p>
<div class="lb-trans"><p>&#34;考虑到人工智能进展的速度，这种能力的传播不会太久，可能超出那些致力于安全部署的行为者。&#34;</p>
</div><h2>Risks highest in legacy systems</h2>
<div class="lb-trans"><h2>遗留系统的风险最高</h2>
</div><p>The potential threat is particularly acute in sectors that rely on complex and ageing infrastructure.</p>
<div class="lb-trans"><p>潜在威胁在依赖复杂和老旧基础设施的行业中尤为严重。</p>
</div><p>A report by Bain &amp; Company noted that industries such as energy, manufacturing, and transportation could face heightened risks due to outdated systems that are difficult to patch.</p>
<div class="lb-trans"><p>贝恩公司的一份报告指出，能源、制造和运输等行业可能因难以修补的过时系统而面临更高的风险。</p>
</div><p>Banks, too, are vulnerable given their reliance on interconnected systems, some of which date back decades.</p>
<div class="lb-trans"><p>银行也面临脆弱性，因为它们依赖于相互连接的系统，其中一些系统可以追溯到几十年前。</p>
</div><p>However, the report emphasised that the risks are manageable with strong cybersecurity practices.</p>
<div class="lb-trans"><p>然而，该报告强调，凭借强大的网络安全实践，这些风险是可控的。</p>
</div><p>&#34;The threat is serious, but it is not insurmountable, and strong cybersecurity foundations are your best defense,&#34; it said, adding that existing safeguards such as network segmentation, access controls, and anomaly detection can provide significant protection.</p>
<div class="lb-trans"><p>&#34;威胁是严重的，但并非不可克服，强大的网络安全基础是你最好的防御，&#34;报告指出，并补充说，现有的保护措施如网络分段、访问控制和异常检测可以提供显著保护。</p>
</div><h2>Focus shifts to preparedness over panic</h2>
<div class="lb-trans"><h2>关注准备而非恐慌</h2>
</div><p>Rather than rushing to develop entirely new AI-specific defences, experts argue that organisations should prioritise strengthening existing security frameworks.</p>
<div class="lb-trans"><p>专家认为，与其急于开发全新的人工智能特定防御，组织应优先加强现有的安全框架。</p>
</div><p>Recommendations include setting up dedicated AI threat response teams, improving baseline cybersecurity measures, and addressing vulnerabilities in operational technology environments.</p>
<div class="lb-trans"><p>建议包括建立专门的人工智能威胁响应团队、改善基础网络安全措施以及解决操作技术环境中的漏洞。</p>
</div><p>As access to powerful AI systems expands, regulators and companies alike face a balancing act—leveraging the benefits of innovation while guarding against unintended consequences.</p>
<div class="lb-trans"><p>随着对强大人工智能系统的访问扩大，监管机构和公司面临着平衡创新收益与防范意外后果的挑战。</p>
</div><p>The rapid global response to Mythos suggests that policymakers are keenly aware of the stakes.</p>
<div class="lb-trans"><p>对 Mythos 的迅速全球反应表明，政策制定者对风险有着敏锐的意识。</p>
</div><p>Whether these early efforts will be enough to mitigate the risks posed by next-generation AI remains an open question.</p>
<div class="lb-trans"><p>这些早期的努力是否足以减轻下一代人工智能带来的风险仍然是一个悬而未决的问题。</p>
</div><p>The post Why Anthropic’s Mythos AI has regulators, central banks on edge appeared first on Invezz</p>
<div class="lb-trans"><p>文章《为什么 Anthropic 的 Mythos AI 让监管机构和中央银行感到紧张》最早出现在 Invezz 上。</p>
</div>

为什么 Anthropic 的 Mythos AI 让监管机构和中央银行感到不安