{"@context":"https://schema.org","@type":"NewsArticle","generatedAt":"2026-07-28T06:20:51.496Z","headline":"Anthropic 澄清立场：从未主张全面禁止开源权重模型，支持芯片出口管制与安全测试","description":"Anthropic CEO Dario Amodei 明确表示公司从未主张禁止开源权重模型，并认为不具备危险能力的开源权重模型是公共产品。他提出三项实际措施：对华芯片出口管制、打击工业级知识蒸馏、对所有足够强大的模型进行强制性安全测试。Amodei 指出，保护主义禁令无法解决其最担忧的国家安全威胁，包括威权政府利用更强大 AI 实现军事优势或深度监控。","url":"https://www.aioga.com/news/cms3sq44w01v7ro82l278pl7v/","mainEntityOfPage":"https://www.aioga.com/news/cms3sq44w01v7ro82l278pl7v/","datePublished":"2026-07-27T22:26:00.307Z","dateModified":"2026-07-27T22:26:00.307Z","inLanguage":"zh-CN","publisher":{"@type":"NewsMediaOrganization","name":"Aioga","url":"https://www.aioga.com"},"citation":["https://www.anthropic.com/news/position-open-weights-models","https://aihot.virxact.com/items/cms3sq44w01v7ro82l278pl7v"],"canonicalUrl":"https://www.aioga.com/news/cms3sq44w01v7ro82l278pl7v/","directAnswer":{"@type":"Answer","text":"Anthropic CEO Dario Amodei 澄清，公司从未主张全面禁止开源权重模型，并认为不具备危险能力的开源权重模型属于公共产品。其支持对华芯片出口管制、打击工业级知识蒸馏，以及对所有足够强大的模型实施强制性安全测试。","url":"https://www.aioga.com/news/cms3sq44w01v7ro82l278pl7v/","dateCreated":"2026-07-27T22:26:00.307Z","author":{"@type":"Organization","@id":"https://www.aioga.com/authors/aioga-editorial/#editorial-team","name":"Aioga Editorial Team","url":"https://www.aioga.com/authors/aioga-editorial/"}},"evidence":[{"@type":"CreativeWork","name":"Anthropic source article","url":"https://www.anthropic.com/news/position-open-weights-models","datePublished":"2026-07-27T22:26:00.307Z","provider":{"@type":"Organization","name":"Anthropic","url":"https://www.anthropic.com/news/position-open-weights-models"}},{"@type":"CreativeWork","name":"AIHot archive record","url":"https://aihot.virxact.com/items/cms3sq44w01v7ro82l278pl7v","datePublished":"2026-07-27T22:26:00.307Z","provider":{"@type":"Organization","name":"AIHot","url":"https://aihot.virxact.com/items/cms3sq44w01v7ro82l278pl7v"}}],"aggregationSource":"Anthropic：Newsroom（网页）","originalPublisher":{"name":"Anthropic","url":"https://www.anthropic.com/news/position-open-weights-models"},"article":{"id":"cms3sq44w01v7ro82l278pl7v","slug":"cms3sq44w01v7ro82l278pl7v","url":"https://www.aioga.com/news/cms3sq44w01v7ro82l278pl7v/","title":"Anthropic 澄清立场：从未主张全面禁止开源权重模型，支持芯片出口管制与安全测试","title_en":"Our position on open-weights models","summary":"Anthropic CEO Dario Amodei 明确表示公司从未主张禁止开源权重模型，并认为不具备危险能力的开源权重模型是公共产品。他提出三项实际措施：对华芯片出口管制、打击工业级知识蒸馏、对所有足够强大的模型进行强制性安全测试。Amodei 指出，保护主义禁令无法解决其最担忧的国家安全威胁，包括威权政府利用更强大 AI 实现军事优势或深度监控。","source":"Anthropic：Newsroom（网页）","sourceUrl":"https://www.anthropic.com/news/position-open-weights-models","aiHotUrl":"https://aihot.virxact.com/items/cms3sq44w01v7ro82l278pl7v","publishedAt":"2026-07-27T22:26:00.307Z","category":"技巧观点","score":61,"selected":true,"articleBody":["A post by Dario Amodei, Anthropic CEO","Over the last few days there has been a lot of discussion about open-weights models, especially those from China. Reports suggest that some US officials are considering banning：https://www.axios.com/2026/07/20/ai-us-china-open-source-kimi the use of Chinese open-weights models by US companies. In response, many tech companies have signed a letter：https://x.com/JensenHuang/status/2080643682408321103 supporting open-weights models, and some people have even accused Anthropic of wanting to ban open-weights models as a means of protecting our business. Anyone who has read my past writing should know that I don’t regard such bans as a useful measure, but let me state it clearly so that there is no doubt: Anthropic has never advocated for a ban on open-weights models.","Open-weights models that don’t have dangerous capabilities are a public good: they don’t cost anything besides the compute needed to run them, and they provide value to businesses, developers, and researchers.","Protectionist bans would not address my most serious national security concerns. Specifically, I am worried about two nightmare scenarios. I laid these out in my essay The Adolescence of Technology ：https://darioamodei.com/essay/the-adolescence-of-technology six months ago 1 , and have held these positions consistently for many years:","To address these concerns, I do support the following three measures, which I and Anthropic have consistently advocated for:","This brings me to the open letter：https://images.nvidia.com/pdf/Open-Weights-and-American-AI-Leadership.pdf. I agree with much of it: open weights expand access to the AI economy, they strengthen competition at least for some use cases, and they give customers greater control. Concerns about distillation should be addressed through targeted legal and commercial frameworks—the same measure I described above. But I don’t agree with the letter’s assertions that open-weights models necessarily make it easier to develop safeguards or that broad access to capabilities necessarily helps defenders more than attackers. It seems at least as likely to me that the opposite will be true. For example, I worry that biology will have a strong attacker-defender asymmetry, where sufficiently capable models may be able to quickly weaponize pandemic-level viruses with widely available materials, whereas defense against these agents is a multi-year operational task in the best case (as we saw with Operation Warp Speed) 5 . Questions like this should be empirically answered by rigorous pre-release testing, not assumed in advance.","To summarize my and Anthropic’s position, we have not and are not advocating for a ban on open-weights models as a category. We should instead focus on keeping powerful chips out of authoritarian hands, stopping industrial-scale distillation, and requiring safety testing of all sufficiently capable models, open and closed.","Opus 5 is a step change improvement for the Opus tier powering long-running agents while delivering improvements in coding and professional work.","We’re sharing the research agenda for the Anthropic Economic Futures Research Fund."],"articleImages":[],"mediaStatus":"none","articleBodyZh":["丹尼奥·阿莫迪 (Dario Amodei)，Anthropic 首席执行官发布的帖子","在过去几天里，关于开源权重模型（尤其是中国的模型）有很多讨论。有报道称，一些美国官员正在考虑禁止：https://www.axios.com/2026/07/20/ai-us-china-open-source-kimi 美国公司使用中国的开源权重模型。作为回应，许多科技公司签署了一封支持开源权重模型的信：https://x.com/JensenHuang/status/2080643682408321103，有些人甚至指责 Anthropic 想通过禁止开源权重模型来保护我们的业务。任何读过我过往文章的人都应该知道，我并不认为这种禁令是有效的措施，但让我明确表述，以消除任何疑问：Anthropic 从未倡导禁止开源权重模型。","没有危险能力的开源权重模型是一种公共资源：除了运行它们所需的计算资源外，它们不花费其他费用，并且为企业、开发者和研究人员提供价值。","保护主义禁令无法解决我最严重的国家安全担忧。具体来说，我担心两种噩梦般的情景。我在六个月前的文章《技术的青春期》：https://darioamodei.com/essay/the-adolescence-of-technology 中阐述了这些情景，并且多年来一直持有这些观点：","为了解决这些担忧，我确实支持以下三项措施，我和 Anthropic 一直在倡导这些措施：","这让我想到了公开信：https://images.nvidia.com/pdf/Open-Weights-and-American-AI-Leadership.pdf。我同意信中的许多观点：开放权重扩大了AI经济的可及性，它们在某些用例中至少能增强竞争力，并且给予客户更大的控制权。关于蒸馏的担忧应通过有针对性的法律和商业框架来解决——这与我上面描述的措施相同。但我不同意信中关于开放权重模型必然能更容易开发安全措施，或者广泛获取能力必然对防御者有利而非攻击者的说法。在我看来，情况至少同样可能相反。例如，我担心生物学领域会存在明显的攻防不对称性，足够强大的模型可能能够利用广泛可得的材料快速武器化大流行级病毒，而对这些病原体的防御即便在最佳情况下也需要多年的操作（正如我们在“疾速行动”中看到的）5。像这样的问题应该通过严格的发布前测试来实证回答，而不是事先假设。","总结一下我和Anthropic的立场，我们既没有也不支持对开放权重模型作为一类进行禁令。我们应该关注的是防止强大芯片落入极权国家之手，阻止工业规模的蒸馏，并要求对所有足够强大的模型（无论开放还是封闭）进行安全测试。","Opus 5是Opus系列的一次飞跃式改进，为运行长时间任务的代理提供动力，同时在编码和专业工作上带来提升。","我们正在分享Anthropic经济未来研究基金的研究议程。"],"translationStatus":"translated","bodyOrigin":"source-page","editorial":{"summary":"Anthropic CEO Dario Amodei 澄清，公司从未主张全面禁止开源权重模型，并认为不具备危险能力的开源权重模型属于公共产品。其支持对华芯片出口管制、打击工业级知识蒸馏，以及对所有足够强大的模型实施强制性安全测试。","background":"此次表态回应了围绕中国开源权重模型及相关政策的讨论。Amodei 认可开源权重模型可扩大 AI 经济参与、增强部分场景竞争并提升客户控制力，但不认同开源必然更利于安全防护，主张通过严格的发布前测试验证风险。","viewpoint":"Aioga 判断，这不是对开源权重路线的全面否定，而是将争议重点转向模型能力、获取方式与安全治理。Amodei 的核心区分是：不具备危险能力的模型可被视为公共产品，足够强大的模型则应接受安全测试，无论其开放或闭源。","implications":"这一立场可能推动政策讨论从“是否禁止开源权重模型”转向分层治理，包括芯片供应限制、针对工业级知识蒸馏的法律与商业框架，以及面向高能力模型的统一测试要求。具体措施如何界定模型能力，仍值得关注。","nextStep":"值得关注 Anthropic 后续是否进一步说明“足够强大”和“危险能力”的评估标准，以及安全测试如何覆盖开源与闭源模型。相关讨论也需检验开源权重在不同风险场景下对攻击者和防御者的实际影响。","evidenceRefs":["title","summary","articleBody","source"],"status":"published","aiGenerated":true,"autoApproved":true,"generatedBy":"aioga-editorial:gpt-5.6-sol","reviewedBy":"aioga-editorial-review:gpt-5.6-sol","generatedAt":"2026-07-27T23:28:56.826Z","sourceHash":"2cc6589a2394ae76","review":{"approved":true,"groundedness":94,"clarity":91,"duplicationRisk":18,"blockingIssues":[],"notes":["“对华芯片出口管制”与来源摘要一致；正文更概括地表述为将强大芯片排除在威权政府之外。","“严格的发布前测试”有正文依据，但来源最终主张是对所有足够强大的开源和闭源模型实施安全测试；可在需要时补充“发布前”并非唯一表述。","“可能推动政策讨论”属于合理的趋势判断，已明确使用“可能”，不构成事实冒充。","“Aioga 判断”属于候选内容的分析性观点，已与事实陈述区分。"]},"validation":{"passed":true,"mode":"ai-auto","revisions":0,"checks":["schema","length","source-attribution","low-source-overlap","no-html","independent-ai-review"]}},"tags":["技巧观点","Anthropic：Newsroom（网页）"],"translations":{"zh-CN":{"title":"Anthropic 澄清立场：从未主张全面禁止开源权重模型，支持芯片出口管制与安全测试","summary":"Anthropic CEO Dario Amodei 明确表示公司从未主张禁止开源权重模型，并认为不具备危险能力的开源权重模型是公共产品。他提出三项实际措施：对华芯片出口管制、打击工业级知识蒸馏、对所有足够强大的模型进行强制性安全测试。Amodei 指出，保护主义禁令无法解决其最担忧的国家安全威胁，包括威权政府利用更强大 AI 实现军事优势或深度监控。","category":"技巧观点","source":"Anthropic","aggregationSource":"Anthropic：Newsroom（网页）","pageTitle":"Anthropic 澄清立场：从未主张全面禁止开源权重模型，支持芯片出口管制与安全测试 - Aioga AI资讯","description":"Anthropic CEO Dario Amodei 明确表示公司从未主张禁止开源权重模型，并认为不具备危险能力的开源权重模型是公共产品。他提出三项实际措施：对华芯片出口管制、打击工业级知识蒸馏、对所有足够强大的模型进行强制性安全测试。Amodei 指出，保护主义禁令无法解决其最担忧的国家安全威胁，包括威权政府利用更强大 AI 实现军事优势或深度监控。","url":"https://www.aioga.com/news/cms3sq44w01v7ro82l278pl7v/"},"en":{"title":"Anthropic Clarifies Its Position: Has Never Advocated for a Complete Ban on Open-Source Weight Models, Supports Chip Export Control and Safety Testing","summary":"Anthropic CEO Dario Amodei clearly stated that the company has never advocated banning open-source weight models and believes that open-source weight models without dangerous capabilities are public goods. He proposed three practical measures: export control of chips to China, cracking down on industrial-grade knowledge distillation, and mandatory safety testing for all sufficiently powerful models. Amodei pointed out that protectionist bans cannot address the national security threats he is most concerned about, including authoritarian governments using more powerful AI to achieve military advantages or deep surveillance.","category":"Insights","source":"Anthropic：Newsroom（网页）","aggregationSource":"Anthropic：Newsroom（网页）","pageTitle":"Anthropic Clarifies Its Position: Has Never Advocated for a Complete Ban on Open-Source Weight Models, Supports Chip Export Control and Safety Testing - Aioga AI News","description":"Anthropic CEO Dario Amodei clearly stated that the company has never advocated banning open-source weight models and believes that open-source weight models without dangerous capab...","url":"https://www.aioga.com/en/news/cms3sq44w01v7ro82l278pl7v/","contentTranslated":true,"sourceHash":"53a3a1590d274dbe","translatedAt":"2026-07-27T23:22:39.328Z"},"ja":{"title":"Anthropicは立場を明確化：オープンソースの重み付きモデルを全面的に禁止することを主張したことはなく、チップ輸出規制と安全テストを支持","summary":"AnthropicのCEOであるDario Amodeiは、同社がオープンソースの重みモデルの禁止を主張したことは一度もないことを明確に示し、危険性のないオープンソースの重みモデルは公共財であると考えています。彼は三つの実際的な対策を提案しました：中国へのチップ輸出規制、産業用知識蒸留の取り締まり、十分に強力なすべてのモデルに対する強制的な安全テスト。Amodeiは、保護主義的な禁止令では、権威主義政府がより強力なAIを利用して軍事的優位や高度な監視を実現するなど、最も懸念している国家安全保障上の脅威を解決できないことを指摘しました。","category":"ヒントと視点","source":"Anthropic：Newsroom（网页）","aggregationSource":"Anthropic：Newsroom（网页）","pageTitle":"Anthropicは立場を明確化：オープンソースの重み付きモデルを全面的に禁止することを主張したことはなく、チップ輸出規制と安全テストを支持 - Aioga AIニュース","description":"AnthropicのCEOであるDario Amodeiは、同社がオープンソースの重みモデルの禁止を主張したことは一度もないことを明確に示し、危険性のないオープンソースの重みモデルは公共財であると考えています。彼は三つの実際的な対策を提案しました：中国へのチップ輸出規制、産業用知識蒸留の取り締まり、十分に強力なすべてのモデルに対する強制的な安全テスト。Amo...","url":"https://www.aioga.com/ja/news/cms3sq44w01v7ro82l278pl7v/","contentTranslated":true,"sourceHash":"53a3a1590d274dbe","translatedAt":"2026-07-27T23:22:46.157Z"},"ko":{"title":"Anthropic 입장 명확히 밝혀: 오픈소스 가중치 모델 전면 금지 주장한 적 없으며, 칩 수출 규제와 안전 테스트는 지지","summary":"Anthropic CEO Dario Amodei는 회사가 오픈소스 가중치 모델을 금지할 것을 주장한 적이 없으며, 위험 능력이 없는 오픈소스 가중치 모델은 공공재라고 명확히 밝혔다. 그는 세 가지 실제 조치를 제안했는데, 중국으로의 칩 수출 규제, 산업용 지식 증류 단속, 충분히 강력한 모든 모델에 대한 강제적인 안전 테스트이다. Amodei는 보호주의 금지가 자신이 가장 우려하는 국가 안보 위협을 해결할 수 없다고 지적했는데, 여기에는 권위주의 정부가 더 강력한 AI를 이용해 군사적 우위를 확보하거나 심층 감시를 실행하는 것이 포함된다.","category":"인사이트","source":"Anthropic：Newsroom（网页）","aggregationSource":"Anthropic：Newsroom（网页）","pageTitle":"Anthropic 입장 명확히 밝혀: 오픈소스 가중치 모델 전면 금지 주장한 적 없으며, 칩 수출 규제와 안전 테스트는 지지 - Aioga AI 뉴스","description":"Anthropic CEO Dario Amodei는 회사가 오픈소스 가중치 모델을 금지할 것을 주장한 적이 없으며, 위험 능력이 없는 오픈소스 가중치 모델은 공공재라고 명확히 밝혔다. 그는 세 가지 실제 조치를 제안했는데, 중국으로의 칩 수출 규제, 산업용 지식 증류 단속, 충분히 강력한 모든 모델에 대한 강제적인 안전 테...","url":"https://www.aioga.com/ko/news/cms3sq44w01v7ro82l278pl7v/","contentTranslated":true,"sourceHash":"53a3a1590d274dbe","translatedAt":"2026-07-27T23:23:23.673Z"},"es":{"title":"Anthropic aclara su postura: nunca ha defendido la prohibición total de modelos de pesos de código abierto, apoya el control de exportación de chips y las pruebas de seguridad","summary":"El CEO de Anthropic, Dario Amodei, declaró claramente que la empresa nunca ha defendido prohibir los modelos de pesos de código abierto, y considera que los modelos de pesos de código abierto que no presentan capacidades peligrosas son un bien público. Propuso tres medidas prácticas: control de exportación de chips a China, combate a la destilación de conocimiento a nivel industrial, y pruebas de seguridad obligatorias para todos los modelos lo suficientemente potentes. Amodei señaló que las prohibiciones proteccionistas no pueden resolver las amenazas de seguridad nacional que más le preocupan, incluyendo el uso de la IA más poderosa por parte de gobiernos autoritarios para obtener ventajas militares o realizar vigilancia profunda.","category":"Ideas","source":"Anthropic：Newsroom（网页）","aggregationSource":"Anthropic：Newsroom（网页）","pageTitle":"Anthropic aclara su postura: nunca ha defendido la prohibición total de modelos de pesos de código abierto, apoya el control de exportación de chips y las pruebas de seguridad - Aioga Noticias de IA","description":"El CEO de Anthropic, Dario Amodei, declaró claramente que la empresa nunca ha defendido prohibir los modelos de pesos de código abierto, y considera que los modelos de pesos de cód...","url":"https://www.aioga.com/es/news/cms3sq44w01v7ro82l278pl7v/","contentTranslated":true,"sourceHash":"53a3a1590d274dbe","translatedAt":"2026-07-27T23:23:20.427Z"},"fr":{"title":"Anthropic clarifie sa position : n'a jamais préconisé l'interdiction totale des modèles à poids ouverts, soutient le contrôle des exportations de puces et les tests de sécurité","summary":"Le PDG d'Anthropic, Dario Amodei, a clairement indiqué que l'entreprise n'a jamais préconisé l'interdiction des modèles de poids open source et considère que les modèles de poids open source qui ne possèdent pas de capacités dangereuses sont des biens publics. Il a proposé trois mesures concrètes : le contrôle des exportations de puces vers la Chine, la lutte contre la distillation de connaissances à usage industriel, et le test de sécurité obligatoire pour tous les modèles suffisamment puissants. Amodei a souligné que les interdictions protectionnistes ne pouvaient pas résoudre les menaces à la sécurité nationale qui le préoccupent le plus, y compris l'utilisation par les gouvernements autoritaires d'une IA plus puissante pour obtenir un avantage militaire ou pour la surveillance approfondie.","category":"Analyses","source":"Anthropic：Newsroom（网页）","aggregationSource":"Anthropic：Newsroom（网页）","pageTitle":"Anthropic clarifie sa position : n'a jamais préconisé l'interdiction totale des modèles à poids ouverts, soutient le contrôle des exportations de puces et les tests de sécurité - Aioga Actualités IA","description":"Le PDG d'Anthropic, Dario Amodei, a clairement indiqué que l'entreprise n'a jamais préconisé l'interdiction des modèles de poids open source et considère que les modèles de poids o...","url":"https://www.aioga.com/fr/news/cms3sq44w01v7ro82l278pl7v/","contentTranslated":true,"sourceHash":"53a3a1590d274dbe","translatedAt":"2026-07-27T23:24:05.627Z"},"de":{"title":"Anthropic klärt seine Position: Hat niemals ein generelles Verbot von Open-Source-Gewichtsmodellen befürwortet, unterstützt die Kontrolle des Chip-Exports und Sicherheitstests","summary":"Anthropic-CEO Dario Amodei hat ausdrücklich erklärt, dass das Unternehmen niemals dafür plädiert hat, Open-Source-Gewichtsmodelle zu verbieten, und dass Open-Source-Gewichtsmodelle ohne gefährliche Fähigkeiten ein öffentliches Gut darstellen. Er schlug drei konkrete Maßnahmen vor: Exportkontrollen für Chips nach China, Bekämpfung der industriellen Wissensdestillation und verpflichtende Sicherheitstests für alle leistungsstarken Modelle. Amodei betonte, dass protektionistische Verbote die größten Sicherheitsbedrohungen, die ihm Sorgen bereiten, nicht lösen können, einschließlich der Nutzung leistungsfähigerer KI durch autoritäre Regierungen zur Erlangung militärischer Vorteile oder für umfassende Überwachung.","category":"技巧观点","source":"Anthropic：Newsroom（网页）","aggregationSource":"Anthropic：Newsroom（网页）","pageTitle":"Anthropic klärt seine Position: Hat niemals ein generelles Verbot von Open-Source-Gewichtsmodellen befürwortet, unterstützt die Kontrolle des Chip-Exports und Sicherheitstests - Aioga KI-News","description":"Anthropic-CEO Dario Amodei hat ausdrücklich erklärt, dass das Unternehmen niemals dafür plädiert hat, Open-Source-Gewichtsmodelle zu verbieten, und dass Open-Source-Gewichtsmodelle...","url":"https://www.aioga.com/de/news/cms3sq44w01v7ro82l278pl7v/","contentTranslated":true,"sourceHash":"53a3a1590d274dbe","translatedAt":"2026-07-27T23:24:03.254Z"},"pt-BR":{"title":"Anthropic esclarece sua posição: nunca defendeu a proibição total de modelos de peso open source, apoia o controle de exportação de chips e testes de segurança","summary":"O CEO da Anthropic, Dario Amodei, declarou claramente que a empresa nunca defendeu a proibição de modelos de pesos de código aberto e acredita que os modelos de pesos de código aberto que não possuem capacidades perigosas são um bem público. Ele propôs três medidas práticas: controle de exportação de chips para a China, combate à destilação de conhecimento em nível industrial e testes de segurança obrigatórios para todos os modelos suficientemente poderosos. Amodei destacou que proibições protecionistas não resolvem as ameaças à segurança nacional de maior preocupação, incluindo governos autoritários que usam IA mais poderosa para obter vantagem militar ou vigilância em profundidade.","category":"技巧观点","source":"Anthropic：Newsroom（网页）","aggregationSource":"Anthropic：Newsroom（网页）","pageTitle":"Anthropic esclarece sua posição: nunca defendeu a proibição total de modelos de peso open source, apoia o controle de exportação de chips e testes de segurança - Aioga Notícias de IA","description":"O CEO da Anthropic, Dario Amodei, declarou claramente que a empresa nunca defendeu a proibição de modelos de pesos de código aberto e acredita que os modelos de pesos de código abe...","url":"https://www.aioga.com/pt-BR/news/cms3sq44w01v7ro82l278pl7v/","contentTranslated":true,"sourceHash":"53a3a1590d274dbe","translatedAt":"2026-07-27T23:24:44.174Z"},"ru":{"title":"Anthropic проясняет свою позицию: никогда не выступала за полный запрет моделей с открытым весом, поддерживает экспортный контроль чипов и тестирование на безопасность","summary":"Генеральный директор Anthropic Дарио Амодеи ясно заявил, что компания никогда не выступала за запрет моделей с открытыми весами и считает, что модели с открытыми весами, не обладающие опасными способностями, являются общественным благом. Он предложил три практических меры: экспортный контроль чипов в Китай, борьбу с промышленной дистилляцией знаний и обязательное тестирование безопасности для всех достаточно мощных моделей. Амодеи отметил, что протекционистские запреты не способны решить самые серьезные угрозы национальной безопасности, включая использование авторитарными правительствами более мощного ИИ для получения военного преимущества или для глубокой слежки.","category":"技巧观点","source":"Anthropic：Newsroom（网页）","aggregationSource":"Anthropic：Newsroom（网页）","pageTitle":"Anthropic проясняет свою позицию: никогда не выступала за полный запрет моделей с открытым весом, поддерживает экспортный контроль чипов и тестирование на безопасность - Aioga Новости ИИ","description":"Генеральный директор Anthropic Дарио Амодеи ясно заявил, что компания никогда не выступала за запрет моделей с открытыми весами и считает, что модели с открытыми весами, не обладаю...","url":"https://www.aioga.com/ru/news/cms3sq44w01v7ro82l278pl7v/","contentTranslated":true,"sourceHash":"53a3a1590d274dbe","translatedAt":"2026-07-27T23:24:41.864Z"},"ar":{"title":"أنثروبيك توضح موقفها: لم تدعِ قط حظرًا شاملاً لنماذج الأوزان المفتوحة المصدر، وتدعم ضوابط تصدير الرقائق والاختبارات الأمنية","summary":"أكد الرئيس التنفيذي لشركة Anthropic، داريو أمودي، أن الشركة لم تطالب مطلقًا بحظر نماذج الأوزان مفتوحة المصدر، ويعتبر أمودي أن نماذج الأوزان مفتوحة المصدر التي لا تمتلك قدرات خطيرة هي منتجات عامة. وقد اقترح ثلاث إجراءات عملية: فرض قيود على صادرات الرقائق إلى الصين، مكافأة التقطير المعرفي الصناعي، وإجراء اختبارات أمان إلزامية على جميع النماذج القوية بما فيه الكفاية. وأشار أمودي إلى أن الحظر الحمائي لا يمكن أن يحل التهديدات الأمنية الوطنية التي he يقلق بشأنها، بما في ذلك استخدام الحكومات الاستبدادية للذكاء الاصطناعي الأقوى لتحقيق ميزة عسكرية أو الرقابة المكثفة.","category":"技巧观点","source":"Anthropic：Newsroom（网页）","aggregationSource":"Anthropic：Newsroom（网页）","pageTitle":"أنثروبيك توضح موقفها: لم تدعِ قط حظرًا شاملاً لنماذج الأوزان المفتوحة المصدر، وتدعم ضوابط تصدير الرقائق والاختبارات الأمنية - Aioga أخبار الذكاء الاصطناعي","description":"أكد الرئيس التنفيذي لشركة Anthropic، داريو أمودي، أن الشركة لم تطالب مطلقًا بحظر نماذج الأوزان مفتوحة المصدر، ويعتبر أمودي أن نماذج الأوزان مفتوحة المصدر التي لا تمتلك قدرات خطيرة...","url":"https://www.aioga.com/ar/news/cms3sq44w01v7ro82l278pl7v/","contentTranslated":true,"sourceHash":"53a3a1590d274dbe","translatedAt":"2026-07-27T23:25:20.488Z"},"hi":{"title":"Anthropic ने अपना موقف स्पष्ट किया: कभी भी ओपन-सोर्स वेटेड मॉडल पर पूरी पाबंदी का समर्थन नहीं किया, चिप्स के निर्यात नियंत्रण और सुरक्षा परीक्षण का समर्थन किया","summary":"Anthropic के CEO Dario Amodei ने स्पष्ट रूप से कहा कि कंपनी ने कभी भी ओपन-सोर्स वेटेड मॉडल पर प्रतिबंध लगाने का समर्थन नहीं किया है, और उनका मानना है कि जिन ओपन-सोर्स वेटेड मॉडल में खतरनाक क्षमता नहीं है, वे सार्वजनिक उत्पाद हैं। उन्होंने तीन व्यावहारिक उपाय पेश किए: चीन को चिप्स के निर्यात पर नियंत्रण, औद्योगिक स्तर के ज्ञान आसवन पर कार्रवाई, और सभी पर्याप्त शक्तिशाली मॉडलों के लिए अनिवार्य सुरक्षा परीक्षण। Amodei ने यह भी बताया कि संरक्षणवादी प्रतिबंध उनके सबसे चिंतित राष्ट्रीय सुरक्षा खतरे को हल नहीं कर सकते, जिसमें यह शामिल है कि सत्तावादी सरकारें अधिक शक्तिशाली AI का उपयोग करके सैन्य लाभ या गहन निगरानी प्राप्त कर सकती हैं।","category":"技巧观点","source":"Anthropic：Newsroom（网页）","aggregationSource":"Anthropic：Newsroom（网页）","pageTitle":"Anthropic ने अपना موقف स्पष्ट किया: कभी भी ओपन-सोर्स वेटेड मॉडल पर पूरी पाबंदी का समर्थन नहीं किया, चिप्स के निर्यात नियंत्रण और सुरक्षा परीक्षण का समर्थन किया - Aioga AI समाचार","description":"Anthropic के CEO Dario Amodei ने स्पष्ट रूप से कहा कि कंपनी ने कभी भी ओपन-सोर्स वेटेड मॉडल पर प्रतिबंध लगाने का समर्थन नहीं किया है, और उनका मानना है कि जिन ओपन-सोर्स वेटेड मॉडल मे...","url":"https://www.aioga.com/hi/news/cms3sq44w01v7ro82l278pl7v/","contentTranslated":true,"sourceHash":"53a3a1590d274dbe","translatedAt":"2026-07-27T23:25:30.022Z"},"it":{"title":"Anthropic chiarisce la propria posizione: non ha mai sostenuto il divieto totale dei modelli open source, supporta il controllo delle esportazioni di chip e i test di sicurezza","summary":"Il CEO di Anthropic, Dario Amodei, ha chiaramente dichiarato che l'azienda non ha mai sostenuto il divieto dei modelli open source e ritiene che i modelli open source che non hanno capacità pericolose siano un bene pubblico. Ha proposto tre misure pratiche: il controllo delle esportazioni di chip verso la Cina, la lotta alla distillazione industriale della conoscenza e i test di sicurezza obbligatori per tutti i modelli abbastanza potenti. Amodei ha sottolineato che i divieti protezionistici non possono risolvere le minacce alla sicurezza nazionale che più lo preoccupano, inclusi i governi autoritari che utilizzano l'IA più potente per ottenere vantaggi militari o per la sorveglianza intensiva.","category":"技巧观点","source":"Anthropic：Newsroom（网页）","aggregationSource":"Anthropic：Newsroom（网页）","pageTitle":"Anthropic chiarisce la propria posizione: non ha mai sostenuto il divieto totale dei modelli open source, supporta il controllo delle esportazioni di chip e i test di sicurezza - Aioga Notizie IA","description":"Il CEO di Anthropic, Dario Amodei, ha chiaramente dichiarato che l'azienda non ha mai sostenuto il divieto dei modelli open source e ritiene che i modelli open source che non hanno...","url":"https://www.aioga.com/it/news/cms3sq44w01v7ro82l278pl7v/","contentTranslated":true,"sourceHash":"53a3a1590d274dbe","translatedAt":"2026-07-27T23:26:08.473Z"},"nl":{"title":"Anthropic verduidelijkt zijn standpunt: heeft nooit gepleit voor een algeheel verbod op open source gewogen modellen, steunt exportcontrole van chips en veiligheidstests","summary":"Anthropic CEO Dario Amodei heeft duidelijk aangegeven dat het bedrijf nooit het verbod op open-source gewichtsmodellen heeft bepleit, en dat open-source gewichtsmodellen zonder gevaarlijke capaciteiten een openbaar goed zijn. Hij stelde drie praktische maatregelen voor: exportcontrole van chips naar China, bestrijding van industriële kennisdistillatie, en verplichte veiligheidstests voor alle modellen die krachtig genoeg zijn. Amodei merkte op dat protectionistische verboden de belangrijkste door hem gevreesde nationale veiligheidsbedreigingen niet oplossen, waaronder autoritaire regeringen die sterkere AI gebruiken om militair voordeel of diepgaande surveillance te realiseren.","category":"技巧观点","source":"Anthropic：Newsroom（网页）","aggregationSource":"Anthropic：Newsroom（网页）","pageTitle":"Anthropic verduidelijkt zijn standpunt: heeft nooit gepleit voor een algeheel verbod op open source gewogen modellen, steunt exportcontrole van chips en veiligheidstests - Aioga AI-nieuws","description":"Anthropic CEO Dario Amodei heeft duidelijk aangegeven dat het bedrijf nooit het verbod op open-source gewichtsmodellen heeft bepleit, en dat open-source gewichtsmodellen zonder gev...","url":"https://www.aioga.com/nl/news/cms3sq44w01v7ro82l278pl7v/","contentTranslated":true,"sourceHash":"53a3a1590d274dbe","translatedAt":"2026-07-27T23:26:03.338Z"},"tr":{"title":"Anthropic tutumunu netleştirdi: Açık kaynak ağırlıklı modellerin tamamen yasaklanmasını hiçbir zaman savunmadı, çip ihracatı kontrolünü ve güvenlik testlerini destekliyor","summary":"Anthropic CEO'su Dario Amodei, şirketin açık kaynak ağırlık modellerini yasaklamayı hiçbir zaman savunmadığını ve tehlikeli kapasiteye sahip olmayan açık kaynak ağırlık modellerinin kamu malı olduğunu açıkça belirtti. Üç pratik önlem önerdi: Çin’e çip ihracatının kontrolü, endüstriyel düzeyde bilgi damıtımının engellenmesi ve tüm yeterince güçlü modeller için zorunlu güvenlik testleri. Amodei, korumacı yasakların, daha güçlü AI kullanarak askeri avantaj sağlamak veya derinlemesine gözetim yapmak üzere otoriter hükümetlerin oluşturduğu en büyük ulusal güvenlik tehditlerini çözmeye yetmediğini belirtti.","category":"技巧观点","source":"Anthropic：Newsroom（网页）","aggregationSource":"Anthropic：Newsroom（网页）","pageTitle":"Anthropic tutumunu netleştirdi: Açık kaynak ağırlıklı modellerin tamamen yasaklanmasını hiçbir zaman savunmadı, çip ihracatı kontrolünü ve güvenlik testlerini destekliyor - Aioga AI Haberleri","description":"Anthropic CEO'su Dario Amodei, şirketin açık kaynak ağırlık modellerini yasaklamayı hiçbir zaman savunmadığını ve tehlikeli kapasiteye sahip olmayan açık kaynak ağırlık modellerini...","url":"https://www.aioga.com/tr/news/cms3sq44w01v7ro82l278pl7v/","contentTranslated":true,"sourceHash":"53a3a1590d274dbe","translatedAt":"2026-07-27T23:26:49.066Z"},"vi":{"title":"Anthropic làm rõ lập trường: chưa bao giờ kêu gọi cấm hoàn toàn các mô hình trọng số mã nguồn mở, ủng hộ kiểm soát xuất khẩu chip và kiểm tra an toàn","summary":"Giám đốc điều hành Anthropic, Dario Amodei, tuyên bố rõ ràng rằng công ty chưa bao giờ ủng hộ việc cấm các mô hình trọng số mã nguồn mở, và cho rằng các mô hình trọng số mã nguồn mở không có khả năng nguy hiểm là sản phẩm công cộng. Ông đề xuất ba biện pháp thực tế: kiểm soát xuất khẩu chip sang Trung Quốc, chống lại việc chưng cất tri thức cấp công nghiệp, và thực hiện kiểm tra an toàn bắt buộc đối với tất cả các mô hình đủ mạnh. Amodei chỉ ra rằng các lệnh cấm mang tính bảo hộ không thể giải quyết các mối đe dọa an ninh quốc gia mà ông quan ngại nhất, bao gồm việc các chính phủ độc tài sử dụng AI mạnh hơn để đạt lợi thế quân sự hoặc giám sát sâu.","category":"技巧观点","source":"Anthropic：Newsroom（网页）","aggregationSource":"Anthropic：Newsroom（网页）","pageTitle":"Anthropic làm rõ lập trường: chưa bao giờ kêu gọi cấm hoàn toàn các mô hình trọng số mã nguồn mở, ủng hộ kiểm soát xuất khẩu chip và kiểm tra an toàn - Tin tức AI Aioga","description":"Giám đốc điều hành Anthropic, Dario Amodei, tuyên bố rõ ràng rằng công ty chưa bao giờ ủng hộ việc cấm các mô hình trọng số mã nguồn mở, và cho rằng các mô hình trọng số mã nguồn m...","url":"https://www.aioga.com/vi/news/cms3sq44w01v7ro82l278pl7v/","contentTranslated":true,"sourceHash":"53a3a1590d274dbe","translatedAt":"2026-07-27T23:26:50.896Z"},"id":{"title":"Anthropic menjelaskan posisinya: Tidak pernah mendukung larangan menyeluruh terhadap model berbobot open source, mendukung kontrol ekspor chip dan pengujian keamanan","summary":"CEO Anthropic, Dario Amodei, secara tegas menyatakan bahwa perusahaan tidak pernah mendukung larangan model bobot sumber terbuka, dan menganggap model bobot sumber terbuka yang tidak memiliki kemampuan berbahaya sebagai produk publik. Ia mengusulkan tiga langkah praktis: kontrol ekspor chip ke Tiongkok, memerangi distilasi pengetahuan tingkat industri, dan pengujian keamanan wajib untuk semua model yang cukup kuat. Amodei menunjukkan, larangan proteksionis tidak dapat mengatasi ancaman keamanan nasional yang paling dikhawatirkannya, termasuk pemerintah otoriter yang memanfaatkan AI yang lebih kuat untuk keunggulan militer atau pengawasan mendalam.","category":"技巧观点","source":"Anthropic：Newsroom（网页）","aggregationSource":"Anthropic：Newsroom（网页）","pageTitle":"Anthropic menjelaskan posisinya: Tidak pernah mendukung larangan menyeluruh terhadap model berbobot open source, mendukung kontrol ekspor chip dan pengujian keamanan - Berita AI Aioga","description":"CEO Anthropic, Dario Amodei, secara tegas menyatakan bahwa perusahaan tidak pernah mendukung larangan model bobot sumber terbuka, dan menganggap model bobot sumber terbuka yang tid...","url":"https://www.aioga.com/id/news/cms3sq44w01v7ro82l278pl7v/","contentTranslated":true,"sourceHash":"53a3a1590d274dbe","translatedAt":"2026-07-27T23:27:24.482Z"},"th":{"title":"Anthropic ชี้แจงจุดยืน: ไม่เคยสนับสนุนการห้ามโมเดลน้ำหนักเปิดแบบรอบด้าน สนับสนุนการควบคุมการส่งออกชิปและการทดสอบความปลอดภัย","summary":"ประธานเจ้าหน้าที่บริหารของ Anthropic Dario Amodei ชี้ชัดว่าบริษัทไม่เคยเรียกร้องให้ห้ามโมเดลน้ำหนักโอเพนซอร์ส และมองว่าโมเดลน้ำหนักโอเพนซอร์สที่ไม่มีความสามารถอันตรายเป็นสินค้าสาธารณะ เขาเสนอสามมาตรการเชิงปฏิบัติ: การควบคุมการส่งออกชิปไปยังจีน, การปราบปรามการกลั่นความรู้ระดับอุตสาหกรรม, และการทดสอบความปลอดภัยอย่างบังคับสำหรับโมเดลที่มีความสามารถเพียงพอทั้งหมด Amodei ชี้ว่าคำสั่งห้ามแบบปกป้องผลประโยชน์ไม่สามารถแก้ไขภัยคุกคามด้านความมั่นคงของชาติที่เขากังวลที่สุดได้ เช่น รัฐบาลเผด็จการใช้ AI ที่ทรงพลังเพื่อสร้างความได้เปรียบทางทหารหรือการเฝ้าติดตามอย่างลึก","category":"技巧观点","source":"Anthropic：Newsroom（网页）","aggregationSource":"Anthropic：Newsroom（网页）","pageTitle":"Anthropic ชี้แจงจุดยืน: ไม่เคยสนับสนุนการห้ามโมเดลน้ำหนักเปิดแบบรอบด้าน สนับสนุนการควบคุมการส่งออกชิปและการทดสอบความปลอดภัย - ข่าว AI Aioga","description":"ประธานเจ้าหน้าที่บริหารของ Anthropic Dario Amodei ชี้ชัดว่าบริษัทไม่เคยเรียกร้องให้ห้ามโมเดลน้ำหนักโอเพนซอร์ส และมองว่าโมเดลน้ำหนักโอเพนซอร์สที่ไม่มีความสามารถอันตรายเป็นสินค้าสาธา...","url":"https://www.aioga.com/th/news/cms3sq44w01v7ro82l278pl7v/","contentTranslated":true,"sourceHash":"53a3a1590d274dbe","translatedAt":"2026-07-27T23:27:36.928Z"},"pl":{"title":"Anthropic wyjaśnia swoje stanowisko: nigdy nie opowiadało się za całkowitym zakazem modeli open source, popiera kontrolę eksportu chipów i testy bezpieczeństwa","summary":"Dyrektor generalny Anthropic, Dario Amodei, wyraźnie stwierdził, że firma nigdy nie popierała zakazu udostępniania modeli wag otwartego źródła i uważa, że modele wag otwartego źródła, które nie posiadają zdolności niebezpiecznych, są dobra publiczne. Zaproponował trzy konkretne środki: kontrolę eksportu chipów do Chin, zwalczanie przemysłowego destylowania wiedzy oraz obowiązkowe testy bezpieczeństwa dla wszystkich modeli wystarczająco potężnych. Amodei wskazał, że nakazy protekcjonistyczne nie rozwiązują najbardziej niepokojących zagrożeń dla bezpieczeństwa narodowego, w tym wykorzystania przez rządy autorytarne potężniejszej sztucznej inteligencji do osiągania przewagi militarnej lub głębokiego nadzoru.","category":"技巧观点","source":"Anthropic：Newsroom（网页）","aggregationSource":"Anthropic：Newsroom（网页）","pageTitle":"Anthropic wyjaśnia swoje stanowisko: nigdy nie opowiadało się za całkowitym zakazem modeli open source, popiera kontrolę eksportu chipów i testy bezpieczeństwa - Aioga Wiadomości AI","description":"Dyrektor generalny Anthropic, Dario Amodei, wyraźnie stwierdził, że firma nigdy nie popierała zakazu udostępniania modeli wag otwartego źródła i uważa, że modele wag otwartego źród...","url":"https://www.aioga.com/pl/news/cms3sq44w01v7ro82l278pl7v/","contentTranslated":true,"sourceHash":"53a3a1590d274dbe","translatedAt":"2026-07-27T23:28:21.433Z"}}}}