{"@context":"https://schema.org","@type":"NewsArticle","generatedAt":"2026-08-26T22:21:03.219Z","headline":"Anthropic 开放 Claude 真实使用数据供外部独立研究，公布试点结果","description":"Anthropic 今年春季启动试点，通过隐私保护工具 Anthropic Insights（原 Clio）向斯坦福大学 SALT Lab、牛津大学人类信息处理实验室及 METR 三个外部机构开放约 25 万段 2026 年 4-5 月的 Claude.ai 或 Claude Code 对话数据，供其独立设计研究并公开发布结果。","url":"https://www.aioga.com/news/cmtaddncd0aq0roj2xmay0scj/","mainEntityOfPage":"https://www.aioga.com/news/cmtaddncd0aq0roj2xmay0scj/","datePublished":"2026-08-26T17:30:30.000Z","dateModified":"2026-08-26T17:30:30.000Z","inLanguage":"zh-CN","publisher":{"@type":"NewsMediaOrganization","name":"Aioga","url":"https://www.aioga.com"},"citation":["https://www.anthropic.com/research/enabling-independent-research","https://aihot.virxact.com/items/cmtaddncd0aq0roj2xmay0scj"],"canonicalUrl":"https://www.aioga.com/news/cmtaddncd0aq0roj2xmay0scj/","directAnswer":{"@type":"Answer","text":"Anthropic 表示已启动外部独立研究试点，通过隐私保护工具 Anthropic Insights 向斯坦福大学 SALT Lab、牛津大学人类信息处理实验室和 METR 开放约 25 万段 Claude.ai 或 Claude Code 对话数据。","url":"https://www.aioga.com/news/cmtaddncd0aq0roj2xmay0scj/","dateCreated":"2026-08-26T17:30:30.000Z","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/research/enabling-independent-research","datePublished":"2026-08-26T17:30:30.000Z","provider":{"@type":"Organization","name":"Anthropic","url":"https://www.anthropic.com/research/enabling-independent-research"}},{"@type":"CreativeWork","name":"AIHot archive record","url":"https://aihot.virxact.com/items/cmtaddncd0aq0roj2xmay0scj","datePublished":"2026-08-26T17:30:30.000Z","provider":{"@type":"Organization","name":"AIHot","url":"https://aihot.virxact.com/items/cmtaddncd0aq0roj2xmay0scj"}}],"aggregationSource":"Anthropic：Research（发表成果 · 网页）","originalPublisher":{"name":"Anthropic","url":"https://www.anthropic.com/research/enabling-independent-research"},"geoDeepAnswer":null,"article":{"id":"cmtaddncd0aq0roj2xmay0scj","slug":"cmtaddncd0aq0roj2xmay0scj","url":"https://www.aioga.com/news/cmtaddncd0aq0roj2xmay0scj/","title":"Anthropic 开放 Claude 真实使用数据供外部独立研究，公布试点结果","title_en":"","summary":"Anthropic 今年春季启动试点，通过隐私保护工具 Anthropic Insights（原 Clio）向斯坦福大学 SALT Lab、牛津大学人类信息处理实验室及 METR 三个外部机构开放约 25 万段 2026 年 4-5 月的 Claude.ai 或 Claude Code 对话数据，供其独立设计研究并公开发布结果。","source":"Anthropic：Research（发表成果 · 网页）","sourceUrl":"https://www.anthropic.com/research/enabling-independent-research","aiHotUrl":"https://aihot.virxact.com/items/cmtaddncd0aq0roj2xmay0scj","publishedAt":"2026-08-26T17:30:30.000Z","category":"行业动态","score":72,"selected":true,"articleBody":["Ensuring the transition to transformative AI goes well requires understanding its impact on people and society. Right now, data on real-world interactions with AI is concentrated in a handful of labs. We think it would be good if more data was made widely available—to researchers, policymakers, and the general public.","Researchers outside the labs have two options. They can draw on analyses the labs publish, which reflect real usage but often answer the lab’s questions, rather than their own. Or they can use public datasets, which they can study however they like, but skew toward more casual use, and may not reflect how most people actually use AI. Neither is sufficient for independent research on how AI is actually being used.","We believe this is the first time external researchers have run public independent studies on an AI company's own usage data. Below, we discuss what the external teams found, what we learned running the pilot, and what we are weighing as we decide how to expand the program more widely. We are also publicly releasing the aggregate data from each project：https://huggingface.co/datasets/Anthropic/enabling-independent-research.","The Social and Language Technologies Lab studied how humans collaborate with AI. They looked at what types of work people bring to AI, what roles humans retain in completing that work, and where human-AI collaboration breaks down. They found:","Read their full writeup here：https://www.alphaxiv.org/abs/2608.human-ai-collaboration-at-scalev1.","The Human Information Processing Lab is studying how people feel while using Claude and how that relates to Claude’s behavior. Their early results indicate:","They are still completing their writeup. When it is public, we will add a link to it here.","METR is estimating real-world productivity gains from coding agents and how these increases in productivity change across model generations. Their analysis of Claude Code conversations is still underway, but early results suggest:","It is valuable to pursue the same problem from different perspectives. Some of our partners’ research questions overlapped with work being pursued internally. For example, METR’s proposal was similar to our economics research on “Agentic coding and persistent returns to expertise：https://www.anthropic.com/research/claude-code-expertise.” We found this overlap valuable: it gave external researchers the chance to examine similar data and draw their own conclusions. Whether those align with ours is something we’ll follow as their study continues. We also connected METR with our Economics team and found that this connection improved both research teams’ work.","Research methods that work internally need to adapt for external partners. When using Anthropic Insights, a researcher writes a question such as, “What type of guidance is this person asking for?” and Claude answers it for every conversation in the study. The answers are then aggregated into categories; researchers only see final categories and the percentage of conversations that fall under each one. Because we are relying on Claude’s judgments, the tool is sensitive to a question’s wording; a poorly phrased one can place conversations into categories that misrepresent them. Because no one can read the underlying conversations, these errors are hard to catch.","Maintaining transparency about misuse without enabling it. Some categories in our partners’ Anthropic Insights outputs surfaced violations of our Acceptable Use Policy or Terms of Service—for instance a category of people seeking guidance on a prohibited activity. We think the public should know about misuse of our platform, so we shared most of these violations. The exceptions were categories that described how users got around our safeguards rather than what they attempted. Less than 5% of categories and conversations were affected in each study, and in each case we told researchers which clusters we had altered or removed and why. As a standard practice, when Anthropic Insights surfaces such violations, we share the aggregated data with our Safeguards team for their review. This is also an important process for our work with external researchers moving forward.","Understanding AI’s effects on society is too big a job for AI companies alone. Real oversight needs external researchers asking their own questions of real-world usage data and publishing what they find independently.","Additionally, we thank Miriam Chaum, Ishita Dasgupta, Esin Durmus, Adam Farina, Zoe Hitzig, Jerry Hong, Devin Kuokka, Hendson Lin, Maxim Massenkoff, Peter McCrory, Maryam Quasto, Nitarshan Rajkumar, Amie Rotherham, Divya Siddarth, Taylor Sorensen, Jerome Swannack, Alex Tamkin, Molly Villagra, Scott White, and Charles Yang for their helpful ideas, discussion, feedback, and support.","For their partnership in this program, we thank Vishakh Padmakumar, Yijia Shao, Jennifer Wang, Diyi Yang, and Dora Zhao from the Social and Language Technologies Lab at Stanford, Tsvetomira Dumbalska, Hannah Rose Kirk, and Christopher Summerfield from the Human Information Processing Lab at Oxford, and Joel Becker (now at Anthropic), Daniel Paleka, and Parker Whitfill from METR.","In this post, we share two results that show how Claude can help life scientists increase the pace of their research.","Here, we identify a few examples of behavioral tendencies in current frontier models and show how they can produce unexpected systemic failures, in hopes of starting a conversation about mitigating these risks.","We're sharing a review of the evidence on worker retraining programs, coauthored by independent researcher David Roodman and Anthropic's Maxim Massenkoff."],"articleImages":[{"sourceUrl":"https://www-cdn.anthropic.com/images/4zrzovbb/website/74256b81ca52aca8cf45a0e4e489d4d4d9d98175-1000x1000.svg","alt":"Enabling independent research on how people use Claude","afterParagraph":0,"url":"/media/articles/cmtaddncd0aq0roj2xmay0scj/4435fb96a7f0b473.jpg"}],"mediaStatus":"ok","articleBodyZh":["确保向变革性人工智能的过渡顺利进行，需要了解其对人类和社会的影响。目前，关于与人工智能的真实世界交互的数据集中在少数实验室中。我们认为，如果有更多数据能够广泛提供给研究人员、政策制定者和公众，会是件好事。","实验室外的研究人员有两种选择。他们可以利用实验室发布的分析，这些分析反映了真实使用情况，但通常回答的是实验室自己的问题，而不是他们自己的问题。或者，他们可以使用公共数据集，自由研究，但这些数据倾向于更随意的使用，可能无法反映大多数人实际使用人工智能的方式。这两种情况都不足以进行独立研究人工智能的真实使用情况。","我们相信，这是外部研究人员首次对人工智能公司自身使用数据进行公开独立研究。下面，我们将讨论外部团队的发现、我们在试点过程中学到的经验，以及在决定如何更广泛地扩展该项目时我们所考虑的内容。我们还将公开发布每个项目的汇总数据：https://huggingface.co/datasets/Anthropic/enabling-independent-research。","社会与语言技术实验室研究了人类如何与人工智能协作。他们观察了人们将何种类型的工作交给人工智能、人类在完成这些工作中保留的角色，以及人类与人工智能协作失效的情况。他们发现：","在这里阅读他们的完整文章：https://www.alphaxiv.org/abs/2608.human-ai-collaboration-at-scalev1。","人类信息处理实验室正在研究人们使用Claude时的感受以及这些感受与Claude行为的关系。他们的早期结果表明：","他们仍在完成报告。一旦公开，我们将在这里添加链接。","METR正在评估代码代理在真实世界中的生产力提升，以及这些生产力提升在不同模型迭代中如何变化。他们对Claude Code对话的分析仍在进行中，但早期结果显示：","从不同的角度研究同一个问题是有价值的。我们的一些合作伙伴的研究问题与内部正在进行的工作有重叠。例如，METR 的提案与我们关于“主动编码与专业知识的持续回报：https://www.anthropic.com/research/claude-code-expertise”的经济学研究类似。我们发现这种重叠是有价值的：它为外部研究人员提供了审查类似数据并得出自己结论的机会。这些结论是否与我们的保持一致，将随着他们的研究继续而观察。我们还将 METR 与我们的经济学团队联系起来，发现这种联系改善了两个研究团队的工作。","在内部有效的研究方法需要适应外部合作伙伴。使用 Anthropic Insights 时，研究人员会写一个问题，例如，“这个人正在寻求什么类型的指导？”然后 Claude 会对研究中的每次对话进行回答。答案随后会被汇总到各个类别中；研究人员只能看到最终类别及每个类别下对话的百分比。由于我们依赖 Claude 的判断，该工具对问题的措辞非常敏感；措辞不当的问题可能会将对话归入误导性的类别。由于没有人能够查看底层对话，这些错误很难被发现。","在保持透明度的同时防止滥用。我们合作伙伴的 Anthropic Insights 输出中，有些类别显示违反了我们的可接受使用政策或服务条款——例如，一类人寻求有关被禁止活动的指导。我们认为公众应该了解对我们平台的滥用情况，因此我们分享了大多数此类违规行为。例外情况是那些描述用户如何绕过我们保护措施而非他们尝试的内容的类别。每项研究中，每个类别和对话受影响的比例不到 5%，在每种情况下，我们都会告知研究人员我们修改或删除了哪些集群及原因。作为标准做法，当 Anthropic Insights 发现此类违规时，我们会将汇总数据与我们的护盾团队共享以供审查。这对我们与外部研究人员的未来合作也是一个重要流程。","理解人工智能对社会的影响对于人工智能公司来说是一个太大的任务。真正的监督需要外部研究人员对现实世界的使用数据提出自己的问题，并独立发布他们发现的结果。","此外，我们感谢Miriam Chaum、Ishita Dasgupta、Esin Durmus、Adam Farina、Zoe Hitzig、Jerry Hong、Devin Kuokka、Hendson Lin、Maxim Massenkoff、Peter McCrory、Maryam Quasto、Nitarshan Rajkumar、Amie Rotherham、Divya Siddarth、Taylor Sorensen、Jerome Swannack、Alex Tamkin、Molly Villagra、Scott White和Charles Yang提供的宝贵意见、讨论、反馈和支持。","对于他们在该项目中的合作，我们感谢斯坦福社会与语言技术实验室的Vishakh Padmakumar、邵一佳、王珈雯、杨迪艺和赵多拉，牛津人类信息处理实验室的Tsvetomira Dumbalska、Hannah Rose Kirk和Christopher Summerfield，以及METR的Joel Becker（现任职于Anthropic）、Daniel Paleka和Parker Whitfill。","在本文中，我们分享了两个结果，展示Claude如何帮助生命科学家加快研究进程。","在这里，我们列举了一些当前前沿模型中的行为倾向实例，并展示这些倾向如何可能导致意外的系统性失败，以期引发关于如何减轻这些风险的讨论。","我们分享了一篇关于工人再培训计划证据的综述，由独立研究员David Roodman与Anthropic的Maxim Massenkoff共同撰写。"],"translationStatus":"translated","bodyOrigin":"source-page","editorial":{"summary":"Anthropic 表示已启动外部独立研究试点，通过隐私保护工具 Anthropic Insights 向斯坦福大学 SALT Lab、牛津大学人类信息处理实验室和 METR 开放约 25 万段 Claude.ai 或 Claude Code 对话数据。","background":"Anthropic 认为，现实世界中的人工智能交互数据目前集中在少数实验室。外部研究者若仅依赖实验室分析或公开数据集，可能分别受研究问题或使用场景代表性限制，因此试点尝试提供更接近真实使用的数据来源。","viewpoint":"Aioga 判断，这一试点的核心价值在于把模型公司的真实使用数据纳入外部研究流程，但现有材料只说明了研究主题和部分早期进展，没有提供足够结果来评价研究结论的稳健性。","implications":"值得关注的是，Anthropic 已表示会公开发布各项目的汇总数据，并权衡如何扩大项目。若后续扩大，研究独立性、隐私保护范围和数据代表性将可能成为评估该计划的重要条件。","nextStep":"下一步应等待三个外部团队发布完整研究成果，重点核对其研究方法、数据范围和限制说明；同时关注 Anthropic 是否公布试点经验、扩展安排及汇总数据的具体内容。","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":"2026-08-26T18:19:35.749Z","sourceHash":"642fd4634d1d5870","review":{"approved":true,"groundedness":96,"clarity":91,"duplicationRisk":8,"blockingIssues":[],"notes":["“研究独立性、隐私保护范围和数据代表性将可能成为评估该计划的重要条件”属于基于材料的合理分析，已通过“可能”明确为判断，不构成事实错误。","“Aioga 判断”明确标注为观点，未将其冒充为 Anthropic 或外部研究团队的结论。"]},"validation":{"passed":true,"mode":"ai-auto","revisions":0,"checks":["schema","length","source-attribution","low-source-overlap","no-html","independent-ai-review"]}},"tags":["行业动态","Anthropic：Research（发表成果 · 网页）"],"translations":{"zh-CN":{"title":"Anthropic 开放 Claude 真实使用数据供外部独立研究，公布试点结果","summary":"Anthropic 今年春季启动试点，通过隐私保护工具 Anthropic Insights（原 Clio）向斯坦福大学 SALT Lab、牛津大学人类信息处理实验室及 METR 三个外部机构开放约 25 万段 2026 年 4-5 月的 Claude.ai 或 Claude Code 对话数据，供其独立设计研究并公开发布结果。","category":"行业动态","source":"Anthropic","aggregationSource":"Anthropic：Research（发表成果 · 网页）","pageTitle":"Anthropic 开放 Claude 真实使用数据供外部独立研究，公布试点结果 - Aioga AI资讯","description":"Anthropic 今年春季启动试点，通过隐私保护工具 Anthropic Insights（原 Clio）向斯坦福大学 SALT Lab、牛津大学人类信息处理实验室及 METR 三个外部机构开放约 25 万段 2026 年 4-5 月的 Claude.ai 或 Claude Code 对话数据，供其独立设计研究并公开发布结果。","url":"https://www.aioga.com/news/cmtaddncd0aq0roj2xmay0scj/","articleBody":["确保向变革性人工智能的过渡顺利进行，需要了解其对人类和社会的影响。目前，关于与人工智能的真实世界交互的数据集中在少数实验室中。我们认为，如果有更多数据能够广泛提供给研究人员、政策制定者和公众，会是件好事。","实验室外的研究人员有两种选择。他们可以利用实验室发布的分析，这些分析反映了真实使用情况，但通常回答的是实验室自己的问题，而不是他们自己的问题。或者，他们可以使用公共数据集，自由研究，但这些数据倾向于更随意的使用，可能无法反映大多数人实际使用人工智能的方式。这两种情况都不足以进行独立研究人工智能的真实使用情况。","我们相信，这是外部研究人员首次对人工智能公司自身使用数据进行公开独立研究。下面，我们将讨论外部团队的发现、我们在试点过程中学到的经验，以及在决定如何更广泛地扩展该项目时我们所考虑的内容。我们还将公开发布每个项目的汇总数据：https://huggingface.co/datasets/Anthropic/enabling-independent-research。","社会与语言技术实验室研究了人类如何与人工智能协作。他们观察了人们将何种类型的工作交给人工智能、人类在完成这些工作中保留的角色，以及人类与人工智能协作失效的情况。他们发现：","在这里阅读他们的完整文章：https://www.alphaxiv.org/abs/2608.human-ai-collaboration-at-scalev1。","人类信息处理实验室正在研究人们使用Claude时的感受以及这些感受与Claude行为的关系。他们的早期结果表明：","他们仍在完成报告。一旦公开，我们将在这里添加链接。","METR正在评估代码代理在真实世界中的生产力提升，以及这些生产力提升在不同模型迭代中如何变化。他们对Claude Code对话的分析仍在进行中，但早期结果显示：","从不同的角度研究同一个问题是有价值的。我们的一些合作伙伴的研究问题与内部正在进行的工作有重叠。例如，METR 的提案与我们关于“主动编码与专业知识的持续回报：https://www.anthropic.com/research/claude-code-expertise”的经济学研究类似。我们发现这种重叠是有价值的：它为外部研究人员提供了审查类似数据并得出自己结论的机会。这些结论是否与我们的保持一致，将随着他们的研究继续而观察。我们还将 METR 与我们的经济学团队联系起来，发现这种联系改善了两个研究团队的工作。","在内部有效的研究方法需要适应外部合作伙伴。使用 Anthropic Insights 时，研究人员会写一个问题，例如，“这个人正在寻求什么类型的指导？”然后 Claude 会对研究中的每次对话进行回答。答案随后会被汇总到各个类别中；研究人员只能看到最终类别及每个类别下对话的百分比。由于我们依赖 Claude 的判断，该工具对问题的措辞非常敏感；措辞不当的问题可能会将对话归入误导性的类别。由于没有人能够查看底层对话，这些错误很难被发现。","在保持透明度的同时防止滥用。我们合作伙伴的 Anthropic Insights 输出中，有些类别显示违反了我们的可接受使用政策或服务条款——例如，一类人寻求有关被禁止活动的指导。我们认为公众应该了解对我们平台的滥用情况，因此我们分享了大多数此类违规行为。例外情况是那些描述用户如何绕过我们保护措施而非他们尝试的内容的类别。每项研究中，每个类别和对话受影响的比例不到 5%，在每种情况下，我们都会告知研究人员我们修改或删除了哪些集群及原因。作为标准做法，当 Anthropic Insights 发现此类违规时，我们会将汇总数据与我们的护盾团队共享以供审查。这对我们与外部研究人员的未来合作也是一个重要流程。","理解人工智能对社会的影响对于人工智能公司来说是一个太大的任务。真正的监督需要外部研究人员对现实世界的使用数据提出自己的问题，并独立发布他们发现的结果。","此外，我们感谢Miriam Chaum、Ishita Dasgupta、Esin Durmus、Adam Farina、Zoe Hitzig、Jerry Hong、Devin Kuokka、Hendson Lin、Maxim Massenkoff、Peter McCrory、Maryam Quasto、Nitarshan Rajkumar、Amie Rotherham、Divya Siddarth、Taylor Sorensen、Jerome Swannack、Alex Tamkin、Molly Villagra、Scott White和Charles Yang提供的宝贵意见、讨论、反馈和支持。","对于他们在该项目中的合作，我们感谢斯坦福社会与语言技术实验室的Vishakh Padmakumar、邵一佳、王珈雯、杨迪艺和赵多拉，牛津人类信息处理实验室的Tsvetomira Dumbalska、Hannah Rose Kirk和Christopher Summerfield，以及METR的Joel Becker（现任职于Anthropic）、Daniel Paleka和Parker Whitfill。","在本文中，我们分享了两个结果，展示Claude如何帮助生命科学家加快研究进程。","在这里，我们列举了一些当前前沿模型中的行为倾向实例，并展示这些倾向如何可能导致意外的系统性失败，以期引发关于如何减轻这些风险的讨论。","我们分享了一篇关于工人再培训计划证据的综述，由独立研究员David Roodman与Anthropic的Maxim Massenkoff共同撰写。"]},"en":{"title":"Anthropic opens Claude's real usage data for external independent research and releases pilot results","summary":"This spring, Anthropic launched a pilot program, opening approximately 250,000 segments of Claude.ai or Claude Code conversation data from April to May 2026 to three external institutions—Stanford University's SALT Lab, the University of Oxford's Human Information Processing Lab, and METR—through the privacy protection tool Anthropic Insights (formerly Clio), for them to independently design research and publicly release the results.","category":"Industry","source":"Anthropic：Research（发表成果 · 网页）","aggregationSource":"Anthropic：Research（发表成果 · 网页）","pageTitle":"Anthropic opens Claude's real usage data for external independent research and releases pilot results - Aioga AI News","description":"This spring, Anthropic launched a pilot program, opening approximately 250,000 segments of Claude.ai or Claude Code conversation data from April to May 2026 to three external insti...","url":"https://www.aioga.com/en/news/cmtaddncd0aq0roj2xmay0scj/","contentTranslated":true,"sourceHash":"b9610319e7bb011d","translatedAt":"2026-08-26T18:02:22.439Z"},"ja":{"title":"Anthropic は Claude の実際の使用データを外部の独立研究向けに公開し、パイロット結果を発表","summary":"Anthropicは今年の春、パイロットを開始し、プライバシー保護ツールAnthropic Insights（旧Clio）を通じて、スタンフォード大学SALT Lab、オックスフォード大学ヒューマンインフォメーションプロセッシングラボ、METRの3つの外部機関に対し、約25万件の2026年4～5月のClaude.aiまたはClaude Codeの対話データを公開し、独自に研究を設計して結果を公開することを可能にした。","category":"業界動向","source":"Anthropic：Research（发表成果 · 网页）","aggregationSource":"Anthropic：Research（发表成果 · 网页）","pageTitle":"Anthropic は Claude の実際の使用データを外部の独立研究向けに公開し、パイロット結果を発表 - Aioga AIニュース","description":"Anthropicは今年の春、パイロットを開始し、プライバシー保護ツールAnthropic Insights（旧Clio）を通じて、スタンフォード大学SALT Lab、オックスフォード大学ヒューマンインフォメーションプロセッシングラボ、METRの3つの外部機関に対し、約25万件の2026年4～5月のClaude.aiまたはClaude Codeの対話データを...","url":"https://www.aioga.com/ja/news/cmtaddncd0aq0roj2xmay0scj/","contentTranslated":true,"sourceHash":"b9610319e7bb011d","translatedAt":"2026-08-26T18:02:33.481Z"},"ko":{"title":"Anthropic는 Claude 실제 사용 데이터를 외부 독립 연구용으로 개방하고, 시범 결과를 공개했습니다","summary":"Anthropic은 올해 봄 파일럿을 시작하여, 개인정보 보호 도구 Anthropic Insights(이전 Clio)를 통해 스탠포드 대학 SALT Lab, 옥스퍼드 대학 인간 정보 처리 실험실, METR 등 세 개의 외부 기관에 2026년 4-5월의 Claude.ai 또는 Claude Code 대화 데이터 약 25만 건을 개방하여, 독립적으로 연구를 설계하고 결과를 공개할 수 있도록 했습니다.","category":"업계 동향","source":"Anthropic：Research（发表成果 · 网页）","aggregationSource":"Anthropic：Research（发表成果 · 网页）","pageTitle":"Anthropic는 Claude 실제 사용 데이터를 외부 독립 연구용으로 개방하고, 시범 결과를 공개했습니다 - Aioga AI 뉴스","description":"Anthropic은 올해 봄 파일럿을 시작하여, 개인정보 보호 도구 Anthropic Insights(이전 Clio)를 통해 스탠포드 대학 SALT Lab, 옥스퍼드 대학 인간 정보 처리 실험실, METR 등 세 개의 외부 기관에 2026년 4-5월의 Claude.ai 또는 Claude Code 대화 데이터 약 25만 건...","url":"https://www.aioga.com/ko/news/cmtaddncd0aq0roj2xmay0scj/","contentTranslated":true,"sourceHash":"b9610319e7bb011d","translatedAt":"2026-08-26T18:03:29.319Z"},"es":{"title":"Anthropic abre los datos de uso real de Claude para investigaciones externas independientes y publica los resultados piloto","summary":"Anthropic lanzó un proyecto piloto esta primavera, abriendo aproximadamente 250,000 fragmentos de datos de conversación de Claude.ai o Claude Code de abril a mayo de 2026 a tres instituciones externas: el SALT Lab de la Universidad de Stanford, el Laboratorio de Procesamiento de Información Humana de la Universidad de Oxford y METR, a través de la herramienta de protección de privacidad Anthropic Insights (anteriormente Clio), para que puedan diseñar investigaciones de manera independiente y publicar los resultados.","category":"Industria","source":"Anthropic：Research（发表成果 · 网页）","aggregationSource":"Anthropic：Research（发表成果 · 网页）","pageTitle":"Anthropic abre los datos de uso real de Claude para investigaciones externas independientes y publica los resultados piloto - Aioga Noticias de IA","description":"Anthropic lanzó un proyecto piloto esta primavera, abriendo aproximadamente 250,000 fragmentos de datos de conversación de Claude.ai o Claude Code de abril a mayo de 2026 a tres in...","url":"https://www.aioga.com/es/news/cmtaddncd0aq0roj2xmay0scj/","contentTranslated":true,"sourceHash":"b9610319e7bb011d","translatedAt":"2026-08-26T18:03:22.108Z"},"fr":{"title":"Anthropic ouvre les données d'utilisation réelle de Claude pour des recherches externes indépendantes et publie les résultats du projet pilote","summary":"Au printemps de cette année, Anthropic a lancé un projet pilote, en ouvrant environ 250 000 segments de données de conversation Claude.ai ou Claude Code d'avril à mai 2026 à trois institutions externes – le SALT Lab de l'Université de Stanford, le Laboratoire de traitement de l'information humaine de l'Université d'Oxford et METR – via l'outil de protection de la vie privée Anthropic Insights (anciennement Clio), afin qu'elles puissent concevoir des recherches de manière indépendante et publier les résultats.","category":"Industrie","source":"Anthropic：Research（发表成果 · 网页）","aggregationSource":"Anthropic：Research（发表成果 · 网页）","pageTitle":"Anthropic ouvre les données d'utilisation réelle de Claude pour des recherches externes indépendantes et publie les résultats du projet pilote - Aioga Actualités IA","description":"Au printemps de cette année, Anthropic a lancé un projet pilote, en ouvrant environ 250 000 segments de données de conversation Claude.ai ou Claude Code d'avril à mai 2026 à trois...","url":"https://www.aioga.com/fr/news/cmtaddncd0aq0roj2xmay0scj/","contentTranslated":true,"sourceHash":"b9610319e7bb011d","translatedAt":"2026-08-26T18:04:22.724Z"},"de":{"title":"Anthropic öffnet echte Nutzungsdaten von Claude für externe unabhängige Forschung und veröffentlicht die Pilotresultate","summary":"Anthropic startete in diesem Frühjahr ein Pilotprojekt, bei dem es über das datenschutzfreundliche Tool Anthropic Insights (ehemals Clio) etwa 250.000 Abschnitte von Claude.ai- oder Claude Code-Konversationsdaten von April bis Mai 2026 den externen Institutionen SALT Lab der Stanford University, dem Human Information Processing Lab der University of Oxford und METR zur Verfügung stellte, damit diese eigenständig Forschung designen und die Ergebnisse veröffentlichen können.","category":"行业动态","source":"Anthropic：Research（发表成果 · 网页）","aggregationSource":"Anthropic：Research（发表成果 · 网页）","pageTitle":"Anthropic öffnet echte Nutzungsdaten von Claude für externe unabhängige Forschung und veröffentlicht die Pilotresultate - Aioga KI-News","description":"Anthropic startete in diesem Frühjahr ein Pilotprojekt, bei dem es über das datenschutzfreundliche Tool Anthropic Insights (ehemals Clio) etwa 250.000 Abschnitte von Claude.ai- ode...","url":"https://www.aioga.com/de/news/cmtaddncd0aq0roj2xmay0scj/","contentTranslated":true,"sourceHash":"b9610319e7bb011d","translatedAt":"2026-08-26T18:04:20.771Z"},"pt-BR":{"title":"Anthropic disponibiliza dados reais de uso do Claude para pesquisas externas independentes e publica resultados do piloto","summary":"A Anthropic lançou um projeto piloto nesta primavera, abrindo cerca de 250.000 diálogos do Claude.ai ou Claude Code de abril a maio de 2026 para três instituições externas — o SALT Lab da Universidade de Stanford, o Laboratório de Processamento de Informação Humana da Universidade de Oxford e o METR — por meio da ferramenta de proteção de privacidade Anthropic Insights (anteriormente Clio), para que realizem pesquisas independentes e publiquem os resultados publicamente.","category":"行业动态","source":"Anthropic：Research（发表成果 · 网页）","aggregationSource":"Anthropic：Research（发表成果 · 网页）","pageTitle":"Anthropic disponibiliza dados reais de uso do Claude para pesquisas externas independentes e publica resultados do piloto - Aioga Notícias de IA","description":"A Anthropic lançou um projeto piloto nesta primavera, abrindo cerca de 250.000 diálogos do Claude.ai ou Claude Code de abril a maio de 2026 para três instituições externas — o SALT...","url":"https://www.aioga.com/pt-BR/news/cmtaddncd0aq0roj2xmay0scj/","contentTranslated":true,"sourceHash":"b9610319e7bb011d","translatedAt":"2026-08-26T18:05:13.784Z"},"ru":{"title":"Компания Anthropic открывает реальные данные использования Claude для внешних независимых исследований и публикует результаты пилотного проекта","summary":"Элентропик этой весной запустила пилотный проект, предоставив внешним организациям Stanford SALT Lab, Лаборатории обработки человеческой информации Оксфордского университета и METR около 250 000 диалогов Claude.ai или Claude Code за апрель-май 2026 года через инструменты защиты конфиденциальности Anthropic Insights (ранее Clio), для самостоятельного проведения исследований и публичного опубликования результатов.","category":"行业动态","source":"Anthropic：Research（发表成果 · 网页）","aggregationSource":"Anthropic：Research（发表成果 · 网页）","pageTitle":"Компания Anthropic открывает реальные данные использования Claude для внешних независимых исследований и публикует результаты пилотного проекта - Aioga Новости ИИ","description":"Элентропик этой весной запустила пилотный проект, предоставив внешним организациям Stanford SALT Lab, Лаборатории обработки человеческой информации Оксфордского университета и METR...","url":"https://www.aioga.com/ru/news/cmtaddncd0aq0roj2xmay0scj/","contentTranslated":true,"sourceHash":"b9610319e7bb011d","translatedAt":"2026-08-26T18:05:12.410Z"},"ar":{"title":"شركة Anthropic تفتح بيانات الاستخدام الحقيقية لـClaude للبحث المستقل الخارجي، وتعلن نتائج المشروع التجريبي","summary":"أطلقت شركة Anthropic هذا الربيع برنامجًا تجريبيًا، من خلال أداة حماية الخصوصية Anthropic Insights (المعروفة سابقًا باسم Clio)، لفتح حوالي 250 ألف مقطع من بيانات محادثات Claude.ai أو Claude Code للفترة بين أبريل ومايو 2026 لثلاث مؤسسات خارجية وهي مختبر SALT بجامعة ستانفورد، ومختبر معالجة المعلومات البشرية بجامعة أكسفورد، وMETR، وذلك ليتمكنوا من تصميم بحوثهم بشكل مستقل ونشر النتائج بشكل علني.","category":"行业动态","source":"Anthropic：Research（发表成果 · 网页）","aggregationSource":"Anthropic：Research（发表成果 · 网页）","pageTitle":"شركة Anthropic تفتح بيانات الاستخدام الحقيقية لـClaude للبحث المستقل الخارجي، وتعلن نتائج المشروع التجريبي - Aioga أخبار الذكاء الاصطناعي","description":"أطلقت شركة Anthropic هذا الربيع برنامجًا تجريبيًا، من خلال أداة حماية الخصوصية Anthropic Insights (المعروفة سابقًا باسم Clio)، لفتح حوالي 250 ألف مقطع من بيانات محادثات Claude.ai أ...","url":"https://www.aioga.com/ar/news/cmtaddncd0aq0roj2xmay0scj/","contentTranslated":true,"sourceHash":"b9610319e7bb011d","translatedAt":"2026-08-26T18:06:12.820Z"},"hi":{"title":"Anthropic ने बाहरी स्वतंत्र शोध के लिए Claude के वास्तविक उपयोग डेटा को खोला, पायलट परिणाम जारी किए","summary":"Anthropic ने इस साल वसंत में पायलट लॉन्च किया, और गोपनीयता सुरक्षा उपकरण Anthropic Insights (पूर्व में Clio) के माध्यम से स्टैनफोर्ड विश्वविद्यालय के SALT लैब, ऑक्सफ़ोर्ड विश्वविद्यालय के मानव सूचना प्रसंस्करण प्रयोगशाला और METR नामक तीन बाहरी संस्थानों को लगभग 2.5 लाख 2026 अप्रैल-मई की Claude.ai या Claude Code संवाद डेटा उपलब्ध कराया, ताकि वे स्वतंत्र रूप से शोध डिज़ाइन कर सकें और परिणाम सार्वजनिक रूप से प्रकाशित कर सकें।","category":"行业动态","source":"Anthropic：Research（发表成果 · 网页）","aggregationSource":"Anthropic：Research（发表成果 · 网页）","pageTitle":"Anthropic ने बाहरी स्वतंत्र शोध के लिए Claude के वास्तविक उपयोग डेटा को खोला, पायलट परिणाम जारी किए - Aioga AI समाचार","description":"Anthropic ने इस साल वसंत में पायलट लॉन्च किया, और गोपनीयता सुरक्षा उपकरण Anthropic Insights (पूर्व में Clio) के माध्यम से स्टैनफोर्ड विश्वविद्यालय के SALT लैब, ऑक्सफ़ोर्ड विश्वविद्...","url":"https://www.aioga.com/hi/news/cmtaddncd0aq0roj2xmay0scj/","contentTranslated":true,"sourceHash":"b9610319e7bb011d","translatedAt":"2026-08-26T18:06:15.769Z"},"it":{"title":"Anthropic apre i dati reali di utilizzo di Claude per la ricerca indipendente esterna, pubblicando i risultati del progetto pilota","summary":"Questo primavera, Anthropic ha avviato un progetto pilota, rendendo disponibili circa 250.000 segmenti di dati di conversazione Claude.ai o Claude Code tra aprile e maggio 2026 tramite lo strumento di protezione della privacy Anthropic Insights (precedentemente Clio) a tre istituzioni esterne: il SALT Lab della Stanford University, il Laboratorio di elaborazione delle informazioni umane dell'Università di Oxford e METR, affinché possano condurre ricerche indipendenti e pubblicare i risultati.","category":"行业动态","source":"Anthropic：Research（发表成果 · 网页）","aggregationSource":"Anthropic：Research（发表成果 · 网页）","pageTitle":"Anthropic apre i dati reali di utilizzo di Claude per la ricerca indipendente esterna, pubblicando i risultati del progetto pilota - Aioga Notizie IA","description":"Questo primavera, Anthropic ha avviato un progetto pilota, rendendo disponibili circa 250.000 segmenti di dati di conversazione Claude.ai o Claude Code tra aprile e maggio 2026 tra...","url":"https://www.aioga.com/it/news/cmtaddncd0aq0roj2xmay0scj/","contentTranslated":true,"sourceHash":"b9610319e7bb011d","translatedAt":"2026-08-26T18:07:07.314Z"},"nl":{"title":"Anthropic opent echte gebruiksgegevens van Claude voor externe onafhankelijk onderzoek en publiceert de resultaten van de pilot","summary":"Anthropic is dit voorjaar gestart met een pilootproject, waarbij ze via de privacybeschermingstool Anthropic Insights (voorheen Clio) ongeveer 250.000 gespreksgegevens van Claude.ai of Claude Code van april-mei 2026 beschikbaar stellen aan drie externe instellingen: het SALT Lab van Stanford University, het Human Information Processing Laboratory van de University of Oxford en METR, zodat zij onafhankelijk onderzoek kunnen ontwerpen en de resultaten openbaar kunnen maken.","category":"行业动态","source":"Anthropic：Research（发表成果 · 网页）","aggregationSource":"Anthropic：Research（发表成果 · 网页）","pageTitle":"Anthropic opent echte gebruiksgegevens van Claude voor externe onafhankelijk onderzoek en publiceert de resultaten van de pilot - Aioga AI-nieuws","description":"Anthropic is dit voorjaar gestart met een pilootproject, waarbij ze via de privacybeschermingstool Anthropic Insights (voorheen Clio) ongeveer 250.000 gespreksgegevens van Claude.a...","url":"https://www.aioga.com/nl/news/cmtaddncd0aq0roj2xmay0scj/","contentTranslated":true,"sourceHash":"b9610319e7bb011d","translatedAt":"2026-08-26T18:07:06.505Z"},"tr":{"title":"Anthropic, Claude'un gerçek kullanım verilerini bağımsız dış araştırmalar için açtı, pilot sonuçları açıkladı","summary":"Anthropic bu bahar pilot programını başlattı ve gizlilik koruma aracı Anthropic Insights (eski adıyla Clio) aracılığıyla Stanford Üniversitesi SALT Lab, Oxford Üniversitesi İnsan Bilgi İşleme Laboratuvarı ve METR olmak üzere üç dış kuruluşa 2026 Nisan-Mayıs aylarına ait yaklaşık 250.000 Claude.ai veya Claude Code diyalog verisi açtı, böylece bu veriler bağımsız olarak araştırma tasarlamak ve sonuçları kamuya açıklamak için kullanılabilecek.","category":"行业动态","source":"Anthropic：Research（发表成果 · 网页）","aggregationSource":"Anthropic：Research（发表成果 · 网页）","pageTitle":"Anthropic, Claude'un gerçek kullanım verilerini bağımsız dış araştırmalar için açtı, pilot sonuçları açıkladı - Aioga AI Haberleri","description":"Anthropic bu bahar pilot programını başlattı ve gizlilik koruma aracı Anthropic Insights (eski adıyla Clio) aracılığıyla Stanford Üniversitesi SALT Lab, Oxford Üniversitesi İnsan B...","url":"https://www.aioga.com/tr/news/cmtaddncd0aq0roj2xmay0scj/","contentTranslated":true,"sourceHash":"b9610319e7bb011d","translatedAt":"2026-08-26T18:08:04.702Z"},"vi":{"title":"Anthropic mở dữ liệu sử dụng thực tế của Claude cho nghiên cứu độc lập bên ngoài, công bố kết quả thử nghiệm","summary":"Anthropic mùa xuân năm nay đã khởi động chương trình thí điểm, thông qua công cụ bảo vệ quyền riêng tư Anthropic Insights (trước đây là Clio) mở khoảng 250.000 đoạn dữ liệu đối thoại Claude.ai hoặc Claude Code từ tháng 4-5 năm 2026 cho ba tổ chức bên ngoài là Phòng thí nghiệm SALT của Đại học Stanford, Phòng thí nghiệm Xử lý Thông tin Con người của Đại học Oxford và METR, để họ tự thiết kế nghiên cứu và công bố kết quả.","category":"行业动态","source":"Anthropic：Research（发表成果 · 网页）","aggregationSource":"Anthropic：Research（发表成果 · 网页）","pageTitle":"Anthropic mở dữ liệu sử dụng thực tế của Claude cho nghiên cứu độc lập bên ngoài, công bố kết quả thử nghiệm - Tin tức AI Aioga","description":"Anthropic mùa xuân năm nay đã khởi động chương trình thí điểm, thông qua công cụ bảo vệ quyền riêng tư Anthropic Insights (trước đây là Clio) mở khoảng 250.000 đoạn dữ liệu đối tho...","url":"https://www.aioga.com/vi/news/cmtaddncd0aq0roj2xmay0scj/","contentTranslated":true,"sourceHash":"b9610319e7bb011d","translatedAt":"2026-08-26T18:08:02.091Z"},"id":{"title":"Anthropic membuka data penggunaan nyata Claude untuk penelitian independen eksternal, mempublikasikan hasil percontohan","summary":"Anthropic memulai program percontohan musim semi ini, membuka sekitar 250.000 cuplikan data percakapan Claude.ai atau Claude Code dari April-Mei 2026 kepada tiga lembaga eksternal, yaitu SALT Lab Universitas Stanford, Laboratorium Pemrosesan Informasi Manusia Universitas Oxford, dan METR, melalui alat perlindungan privasi Anthropic Insights (sebelumnya Clio), untuk digunakan dalam desain penelitian independen dan publikasi hasil secara terbuka.","category":"行业动态","source":"Anthropic：Research（发表成果 · 网页）","aggregationSource":"Anthropic：Research（发表成果 · 网页）","pageTitle":"Anthropic membuka data penggunaan nyata Claude untuk penelitian independen eksternal, mempublikasikan hasil percontohan - Berita AI Aioga","description":"Anthropic memulai program percontohan musim semi ini, membuka sekitar 250.000 cuplikan data percakapan Claude.ai atau Claude Code dari April-Mei 2026 kepada tiga lembaga eksternal,...","url":"https://www.aioga.com/id/news/cmtaddncd0aq0roj2xmay0scj/","contentTranslated":true,"sourceHash":"b9610319e7bb011d","translatedAt":"2026-08-26T18:08:49.785Z"},"th":{"title":"Anthropic เปิดเผยข้อมูลการใช้งานจริงของ Claude สำหรับการวิจัยอิสระภายนอก และเผยแพร่ผลการทดลองนำร่อง","summary":"Anthropic ในฤดูใบไม้ผลิปีนี้เริ่มการทดลองนำร่อง โดยเปิดข้อมูลการสนทนา Claude.ai หรือ Claude Code ประมาณ 250,000 ชิ้นช่วงเดือนเมษายน-พฤษภาคม 2026 ให้กับสถาบันภายนอกสามแห่ง ได้แก่ ห้องทดลอง SALT ของมหาวิทยาลัยสแตนฟอร์ด ห้องทดลองประมวลผลข้อมูลมนุษย์ของมหาวิทยาลัยออกซ์ฟอร์ด และ METR ผ่านเครื่องมือปกป้องความเป็นส่วนตัว Anthropic Insights (เดิมชื่อ Clio) เพื่อให้แต่ละสถาบันออกแบบการวิจัยด้วยตนเองและเผยแพร่ผลลัพธ์อย่างเปิดเผย","category":"行业动态","source":"Anthropic：Research（发表成果 · 网页）","aggregationSource":"Anthropic：Research（发表成果 · 网页）","pageTitle":"Anthropic เปิดเผยข้อมูลการใช้งานจริงของ Claude สำหรับการวิจัยอิสระภายนอก และเผยแพร่ผลการทดลองนำร่อง - ข่าว AI Aioga","description":"Anthropic ในฤดูใบไม้ผลิปีนี้เริ่มการทดลองนำร่อง โดยเปิดข้อมูลการสนทนา Claude.ai หรือ Claude Code ประมาณ 250,000 ชิ้นช่วงเดือนเมษายน-พฤษภาคม 2026 ให้กับสถาบันภายนอกสามแห่ง ได้แก่ ห้...","url":"https://www.aioga.com/th/news/cmtaddncd0aq0roj2xmay0scj/","contentTranslated":true,"sourceHash":"b9610319e7bb011d","translatedAt":"2026-08-26T18:09:07.946Z"},"pl":{"title":"Anthropic udostępnia dane rzeczywistego użycia Claude do niezależnych badań zewnętrznych, publikując wyniki pilotażowe","summary":"Anthropic na wiosnę tego roku uruchomił pilotaż, udostępniając za pomocą narzędzia do ochrony prywatności Anthropic Insights (wcześniej Clio) około 250 000 fragmentów danych rozmów Claude.ai lub Claude Code z kwietnia-maja 2026 r. trzem zewnętrznym instytucjom: SALT Lab Uniwersytetu Stanforda, Laboratorium Przetwarzania Informacji Ludzkiej Uniwersytetu Oksfordzkiego oraz METR, w celu samodzielnego prowadzenia badań i publicznego udostępniania wyników.","category":"行业动态","source":"Anthropic：Research（发表成果 · 网页）","aggregationSource":"Anthropic：Research（发表成果 · 网页）","pageTitle":"Anthropic udostępnia dane rzeczywistego użycia Claude do niezależnych badań zewnętrznych, publikując wyniki pilotażowe - Aioga Wiadomości AI","description":"Anthropic na wiosnę tego roku uruchomił pilotaż, udostępniając za pomocą narzędzia do ochrony prywatności Anthropic Insights (wcześniej Clio) około 250 000 fragmentów danych rozmów...","url":"https://www.aioga.com/pl/news/cmtaddncd0aq0roj2xmay0scj/","contentTranslated":true,"sourceHash":"b9610319e7bb011d","translatedAt":"2026-08-26T18:10:16.651Z"}},"evidenceTier":"verified-news","reviewStatus":"editorial-selected","indexable":true,"editorialCover":""}}