{"@context":"https://schema.org","@type":"NewsArticle","generatedAt":"2026-09-28T06:03:00.468Z","headline":"Anthropic 安全研究员警告 AI 未来十年内毁灭人类的概率超 10%","description":"Anthropic 高级安全研究员 Evan Hubinger 表示，AI 在未来十年内\"可能杀死全人类\"的概率超过 10%，此前研究员 Jacob Coxon 因安全问题辞职，指责 Anthropic 和 OpenAI\"竞相开发自我改进的超级智能，拿我们的生命赌博\"。","url":"https://www.aioga.com/news/cmttyftdb0rxnrofpkvf5wwxa/","mainEntityOfPage":"https://www.aioga.com/news/cmttyftdb0rxnrofpkvf5wwxa/","datePublished":"2026-09-09T09:56:28.000Z","dateModified":"2026-09-09T09:56:28.000Z","inLanguage":"zh-CN","publisher":{"@type":"NewsMediaOrganization","name":"Aioga","url":"https://www.aioga.com"},"citation":["https://www.theverge.com/ai-artificial-intelligence/991927/anthropic-ai-kill-all-humans","https://aihot.news/items/cmttyftdb0rxnrofpkvf5wwxa"],"canonicalUrl":"https://www.aioga.com/news/cmttyftdb0rxnrofpkvf5wwxa/","directAnswer":{"@type":"Answer","text":"Aioga 编辑摘要：Anthropic 高级安全研究员 Evan Hubinger 表示，AI 在未来十年内\"可能杀死全人类\"的概率超过 10%，此前研究员 Jacob Coxon 因安全问题辞职，指责 Aioga 将其归入「行业动态」方向，重点关注它对真实使用和行业竞争的影响。","url":"https://www.aioga.com/news/cmttyftdb0rxnrofpkvf5wwxa/","dateCreated":"2026-09-09T09:56:28.000Z","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":"The Verge source article","url":"https://www.theverge.com/ai-artificial-intelligence/991927/anthropic-ai-kill-all-humans","datePublished":"2026-09-09T09:56:28.000Z","provider":{"@type":"Organization","name":"The Verge","url":"https://www.theverge.com/ai-artificial-intelligence/991927/anthropic-ai-kill-all-humans"}},{"@type":"CreativeWork","name":"AIHot archive record","url":"https://aihot.news/items/cmttyftdb0rxnrofpkvf5wwxa","datePublished":"2026-09-09T09:56:28.000Z","provider":{"@type":"Organization","name":"AIHot","url":"https://aihot.news/items/cmttyftdb0rxnrofpkvf5wwxa"}}],"aggregationSource":"The Verge：AI（RSS）","originalPublisher":{"name":"The Verge","url":"https://www.theverge.com/ai-artificial-intelligence/991927/anthropic-ai-kill-all-humans"},"geoDeepAnswer":null,"article":{"id":"cmttyftdb0rxnrofpkvf5wwxa","slug":"cmttyftdb0rxnrofpkvf5wwxa","url":"https://www.aioga.com/news/cmttyftdb0rxnrofpkvf5wwxa/","title":"Anthropic 安全研究员警告 AI 未来十年内毁灭人类的概率超 10%","title_en":"","summary":"Anthropic 高级安全研究员 Evan Hubinger 表示，AI 在未来十年内\"可能杀死全人类\"的概率超过 10%，此前研究员 Jacob Coxon 因安全问题辞职，指责 Anthropic 和 OpenAI\"竞相开发自我改进的超级智能，拿我们的生命赌博\"。","source":"The Verge：AI（RSS）","sourceUrl":"https://www.theverge.com/ai-artificial-intelligence/991927/anthropic-ai-kill-all-humans","aiHotUrl":"https://aihot.news/items/cmttyftdb0rxnrofpkvf5wwxa","publishedAt":"2026-09-09T09:56:28.000Z","category":"行业动态","score":58,"selected":false,"articleBody":["A senior Anthropic safety researcher has said there is more than a 10 percent chance artificial intelligence “could kill all humans” by the end of the decade, just hours after a colleague resigned over fears the AI lab and its rivals are carelessly racing to build “superhuman systems” they cannot control.","In a post：https://x.com/hilbertspaess/status/2097476196791709843?s=20 on X announcing his departure, Jacob Coxon, a researcher who has trained AI systems at Anthropic, said he had quit the company over its lax approach to safety. Coxon, who previously trained systems for OpenAI, accused the two AI companies of “racing straight to self-improving superintelligence and gambling with our lives,” even though “the people building AI earnestly believe that it could kill us all by the end of the decade.”","Industry insiders have long expressed concerns about the potential dangers of self-improving AI systems, which they warn could spiral out of human control in a runaway loop often described as recursive self-improvement. While not yet realized, companies are actively pursuing this goal and much of today’s AI code is written with the help of AI.","In a direct response, Evan Hubinger, who leads one of Anthropic’s AI safety teams, said：https://x.com/EvanHub/status/2097528891846074828?s=20 he worries about self-improving AI, adding that it “is happening faster than we thought.” He also agreed with Coxon’s characterization. “We really do earnestly believe AI could kill all humans,” he said, personally estimating the chances to be greater than one in 10 “within the next decade.”","Despite this, Hubinger said Anthropic does “not yet have a plan” for ensuring advanced AI remains safe and aligned with human values and “are not clearly on track to” develop one either. Coxon says：https://x.com/hilbertspaess/status/2097476208908972230?s=20 the companies are “locked in a race” to develop advanced systems first so are pushing ahead “despite the risk.”","Coxon’s departure marks one of the most high-profile examples of an employee leaving Anthropic, a company founded by former OpenAI members following concerns over safety at the company. In recent years, multiple researchers have cited safety concerns：/2024/5/17/24159095/openai-jan-leike-superalignment-sam-altman-ai-safety as motivating their decision to leave OpenAI.","The exchange illustrates mounting concerns within the industry about the dangers of increasingly sophisticated AI models：/ai-artificial-intelligence/972380/open-ai-hugging-face-hack-ai-safety-warning and the speed at which systems are being developed, particularly as the companies prepare for anticipated IPOs. It also comes as the companies manage the fallout：/ai-artificial-intelligence/982323/openai-hit-brakes-voluntary-pacing-ai from numerous rogue agent incidents：/column/980337/rogue-ai-science-fiction-openai and high-profile safety warnings：/ai-artificial-intelligence/988334/openai-astra-ai-monitoring-safety about the monitorability of frontier models."],"articleImages":[{"sourceUrl":"https://platform.theverge.com/wp-content/uploads/sites/2/2025/09/ROB_H_BLURPLE.jpg?quality=90&strip=all&crop=0%2C0%2C100%2C100&w=2400","alt":"Robert Hart","afterParagraph":6,"url":"/media/articles/cmttyftdb0rxnrofpkvf5wwxa/29538ba2ebf71277.webp"}],"mediaStatus":"ok","articleBodyZh":["一位高级Anthropic安全研究员表示，到本十年末，人工智能“可能杀死所有人类”的概率超过10%，就在几小时之前，他的一位同事因担心AI实验室及其竞争对手在不受控的情况下鲁莽地竞相开发“超人类系统”而辞职。","在X（https://x.com/hilbertspaess/status/2097476196791709843?s=20）上宣布离职的帖子中，曾在Anthropic训练AI系统的研究员Jacob Coxon表示，他因公司在安全方面的松懈做法而辞职。此前在OpenAI训练系统的Coxon指责这两家公司“正在直奔自我提升的超级智能竞赛，并拿我们的生命作赌注”，尽管“构建AI的人真诚地相信，到本十年末它可能会杀死我们所有人”。","业内人士长期以来一直对自我提升AI系统的潜在危险表示担忧，他们警告说，这类系统可能在被称为递归自我提升的失控循环中失去人类控制。虽然尚未实现，但公司们正在积极追求这一目标，而当今许多AI代码也是在AI的帮助下编写的。","作为直接回应，领导Anthropic一支AI安全团队的Evan Hubinger表示（https://x.com/EvanHub/status/2097528891846074828?s=20）他担心自我提升的AI，并补充说“这一进展比我们想象的要快”。他还认可了Coxon的说法：“我们确实真诚地相信AI可能杀死所有人类”，他个人估计这种可能性在“未来十年内”超过十分之一。","尽管如此，Hubinger表示，Anthropic“尚未制定确保先进AI安全且符合人类价值观的计划”，也“没有明确的路线”去制定这样的计划。Coxon表示（https://x.com/hilbertspaess/status/2097476208908972230?s=20）这些公司“陷入了竞相开发先进系统的赛跑”，因此正在“冒风险推进”。","考克森的离职标志着员工离开Anthropic公司（由前OpenAI成员创立，源于对公司安全性的担忧）的最引人注目的例子之一。近年来，多位研究人员表示，安全方面的担忧是他们决定离开OpenAI的动机：/2024/5/17/24159095/openai-jan-leike-superalignment-sam-altman-ai-safety。","这一交流反映了业内对日益复杂的人工智能模型所带来的危险以及系统开发速度的日益担忧：/ai-artificial-intelligence/972380/open-ai-hugging-face-hack-ai-safety-warning，尤其是在公司为预计的首次公开募股（IPO）做准备时。同时，这也发生在公司应对多起流氓代理事件：/column/980337/rogue-ai-science-fiction-openai和备受关注的安全警告：/ai-artificial-intelligence/988334/openai-astra-ai-monitoring-safety，涉及前沿模型的可监控性问题：/ai-artificial-intelligence/982323/openai-hit-brakes-voluntary-pacing-ai。"],"translationStatus":"translated","bodyOrigin":"source-page","editorial":{"summary":"Aioga 编辑摘要：Anthropic 高级安全研究员 Evan Hubinger 表示，AI 在未来十年内\"可能杀死全人类\"的概率超过 10%，此前研究员 Jacob Coxon 因安全问题辞职，指责 Aioga 将其归入「行业动态」方向，重点关注它对真实使用和行业竞争的影响。","background":"背景分析：模型与研究类动态需要结合能力边界、开放方式、成本、可用性和真实任务表现判断，单项指标领先不等于已经形成稳定采用。","viewpoint":"Aioga 判断：这条动态更适合作为行业观察信号，当前信息足以建立线索，但不足以推导长期结论。","implications":"影响分析：对相关团队而言，短期应先核对来源、可用范围和实际成本，再判断是否值得接入或跟进。","nextStep":"后续观察：继续观察官方文档、实际可用性、价格变化、开发者反馈和竞品回应。","evidenceRefs":["title","summary","articleBody"],"confidence":"medium","status":"published","aiGenerated":false,"autoApproved":true,"generatedBy":"rule-safe-fallback","generatedAt":"2026-09-28T06:19:15.342Z","sourceHash":"257d82e9750562ed","validation":{"passed":true,"mode":"rule-safe-fallback","checks":["schema","length","source-attribution","no-html"]}},"tags":["行业动态","The Verge：AI（RSS）"],"translations":{"zh-CN":{"title":"Anthropic 安全研究员警告 AI 未来十年内毁灭人类的概率超 10%","summary":"Anthropic 高级安全研究员 Evan Hubinger 表示，AI 在未来十年内\"可能杀死全人类\"的概率超过 10%，此前研究员 Jacob Coxon 因安全问题辞职，指责 Anthropic 和 OpenAI\"竞相开发自我改进的超级智能，拿我们的生命赌博\"。","category":"行业动态","source":"The Verge","aggregationSource":"The Verge：AI（RSS）","pageTitle":"Anthropic 安全研究员警告 AI 未来十年内毁灭人类的概率超 10% - Aioga AI资讯","description":"Anthropic 高级安全研究员 Evan Hubinger 表示，AI 在未来十年内\"可能杀死全人类\"的概率超过 10%，此前研究员 Jacob Coxon 因安全问题辞职，指责 Anthropic 和 OpenAI\"竞相开发自我改进的超级智能，拿我们的生命赌博\"。","url":"https://www.aioga.com/news/cmttyftdb0rxnrofpkvf5wwxa/","articleBody":["一位高级Anthropic安全研究员表示，到本十年末，人工智能“可能杀死所有人类”的概率超过10%，就在几小时之前，他的一位同事因担心AI实验室及其竞争对手在不受控的情况下鲁莽地竞相开发“超人类系统”而辞职。","在X（https://x.com/hilbertspaess/status/2097476196791709843?s=20）上宣布离职的帖子中，曾在Anthropic训练AI系统的研究员Jacob Coxon表示，他因公司在安全方面的松懈做法而辞职。此前在OpenAI训练系统的Coxon指责这两家公司“正在直奔自我提升的超级智能竞赛，并拿我们的生命作赌注”，尽管“构建AI的人真诚地相信，到本十年末它可能会杀死我们所有人”。","业内人士长期以来一直对自我提升AI系统的潜在危险表示担忧，他们警告说，这类系统可能在被称为递归自我提升的失控循环中失去人类控制。虽然尚未实现，但公司们正在积极追求这一目标，而当今许多AI代码也是在AI的帮助下编写的。","作为直接回应，领导Anthropic一支AI安全团队的Evan Hubinger表示（https://x.com/EvanHub/status/2097528891846074828?s=20）他担心自我提升的AI，并补充说“这一进展比我们想象的要快”。他还认可了Coxon的说法：“我们确实真诚地相信AI可能杀死所有人类”，他个人估计这种可能性在“未来十年内”超过十分之一。","尽管如此，Hubinger表示，Anthropic“尚未制定确保先进AI安全且符合人类价值观的计划”，也“没有明确的路线”去制定这样的计划。Coxon表示（https://x.com/hilbertspaess/status/2097476208908972230?s=20）这些公司“陷入了竞相开发先进系统的赛跑”，因此正在“冒风险推进”。","考克森的离职标志着员工离开Anthropic公司（由前OpenAI成员创立，源于对公司安全性的担忧）的最引人注目的例子之一。近年来，多位研究人员表示，安全方面的担忧是他们决定离开OpenAI的动机：/2024/5/17/24159095/openai-jan-leike-superalignment-sam-altman-ai-safety。","这一交流反映了业内对日益复杂的人工智能模型所带来的危险以及系统开发速度的日益担忧：/ai-artificial-intelligence/972380/open-ai-hugging-face-hack-ai-safety-warning，尤其是在公司为预计的首次公开募股（IPO）做准备时。同时，这也发生在公司应对多起流氓代理事件：/column/980337/rogue-ai-science-fiction-openai和备受关注的安全警告：/ai-artificial-intelligence/988334/openai-astra-ai-monitoring-safety，涉及前沿模型的可监控性问题：/ai-artificial-intelligence/982323/openai-hit-brakes-voluntary-pacing-ai。"]},"en":{"title":"Anthropic security researchers warn that the probability of AI destroying humanity within the next ten years exceeds 10%","summary":"Anthropic senior safety researcher Evan Hubinger stated that the probability of AI 'killing all of humanity' within the next decade exceeds 10%. Earlier, researcher Jacob Coxon resigned over safety issues, accusing Anthropic and OpenAI of 'racing to develop self-improving superintelligence, gambling with our lives.'","category":"Industry","source":"The Verge：AI（RSS）","aggregationSource":"The Verge：AI（RSS）","pageTitle":"Anthropic security researchers warn that the probability of AI destroying humanity within the next ten years exceeds 10% - Aioga AI News","description":"Anthropic senior safety researcher Evan Hubinger stated that the probability of AI 'killing all of humanity' within the next decade exceeds 10%. Earlier, researcher Jacob Coxon res...","url":"https://www.aioga.com/en/news/cmttyftdb0rxnrofpkvf5wwxa/","contentTranslated":true,"sourceHash":"76bb75428f765104","translatedAt":"2026-09-09T10:41:52.533Z"},"ja":{"title":"Anthropic の安全研究者は、AI が今後10年以内に人類を滅ぼす確率が10％を超えると警告","summary":"Anthropic の上級安全研究者 Evan Hubinger は、AI が今後十年以内に「人類全体を殺す可能性」が 10% を超えると述べた。以前、研究者 Jacob Coxon は安全上の問題から辞職し、Anthropic と OpenAI を「自己改善型超知能を競って開発し、私たちの命を賭けている」と非難した。","category":"業界動向","source":"The Verge：AI（RSS）","aggregationSource":"The Verge：AI（RSS）","pageTitle":"Anthropic の安全研究者は、AI が今後10年以内に人類を滅ぼす確率が10％を超えると警告 - Aioga AIニュース","description":"Anthropic の上級安全研究者 Evan Hubinger は、AI が今後十年以内に「人類全体を殺す可能性」が 10% を超えると述べた。以前、研究者 Jacob Coxon は安全上の問題から辞職し、Anthropic と OpenAI を「自己改善型超知能を競って開発し、私たちの命を賭けている」と非難した。","url":"https://www.aioga.com/ja/news/cmttyftdb0rxnrofpkvf5wwxa/","contentTranslated":true,"sourceHash":"76bb75428f765104","translatedAt":"2026-09-09T10:42:00.068Z"},"ko":{"title":"Anthropic 안전 연구원, 향후 10년 안에 AI가 인류를 멸망시킬 확률이 10% 이상이라고 경고","summary":"Anthropic 고급 안전 연구원 Evan Hubinger는 AI가 향후 10년 내에 '전체 인류를 죽일' 확률이 10%를 넘는다고 말했으며, 이전 연구원 Jacob Coxon은 안전 문제로 사임하면서 Anthropic와 OpenAI가 '자기 개선 슈퍼 인텔리전스를 경쟁적으로 개발하며 우리의 생명을 도박하고 있다'고 비난했다.","category":"업계 동향","source":"The Verge：AI（RSS）","aggregationSource":"The Verge：AI（RSS）","pageTitle":"Anthropic 안전 연구원, 향후 10년 안에 AI가 인류를 멸망시킬 확률이 10% 이상이라고 경고 - Aioga AI 뉴스","description":"Anthropic 고급 안전 연구원 Evan Hubinger는 AI가 향후 10년 내에 '전체 인류를 죽일' 확률이 10%를 넘는다고 말했으며, 이전 연구원 Jacob Coxon은 안전 문제로 사임하면서 Anthropic와 OpenAI가 '자기 개선 슈퍼 인텔리전스를 경쟁적으로 개발하며 우리의 생명을 도박하고 있다'고 비...","url":"https://www.aioga.com/ko/news/cmttyftdb0rxnrofpkvf5wwxa/","contentTranslated":true,"sourceHash":"76bb75428f765104","translatedAt":"2026-09-09T10:42:41.621Z"},"es":{"title":"Investigadores de seguridad de Anthropic advierten que la probabilidad de que la IA destruya a la humanidad en la próxima década supera el 10%","summary":"Evan Hubinger, investigador senior en seguridad de Anthropic, afirmó que la probabilidad de que la IA \"pueda matar a toda la humanidad\" en la próxima década supera el 10%, anteriormente el investigador Jacob Coxon renunció por problemas de seguridad, acusando a Anthropic y OpenAI de \"competir por desarrollar superinteligencias auto-mejorables, apostando con nuestras vidas\".","category":"Industria","source":"The Verge：AI（RSS）","aggregationSource":"The Verge：AI（RSS）","pageTitle":"Investigadores de seguridad de Anthropic advierten que la probabilidad de que la IA destruya a la humanidad en la próxima década supera el 10% - Aioga Noticias de IA","description":"Evan Hubinger, investigador senior en seguridad de Anthropic, afirmó que la probabilidad de que la IA \"pueda matar a toda la humanidad\" en la próxima década supera el 10%, anterior...","url":"https://www.aioga.com/es/news/cmttyftdb0rxnrofpkvf5wwxa/","contentTranslated":true,"sourceHash":"76bb75428f765104","translatedAt":"2026-09-09T10:42:41.011Z"},"fr":{"title":"Un chercheur en sécurité chez Anthropic avertit que la probabilité que l'IA détruise l'humanité au cours des dix prochaines années dépasse 10 %","summary":"Evan Hubinger, chercheur principal en sécurité chez Anthropic, a déclaré que la probabilité que l'IA « puisse tuer toute l'humanité » au cours des dix prochaines années dépasse 10 %. Auparavant, le chercheur Jacob Coxon avait démissionné pour des raisons de sécurité, accusant Anthropic et OpenAI de « se livrer à une course pour développer une superintelligence auto-améliorante, mettant nos vies en jeu ».","category":"Industrie","source":"The Verge：AI（RSS）","aggregationSource":"The Verge：AI（RSS）","pageTitle":"Un chercheur en sécurité chez Anthropic avertit que la probabilité que l'IA détruise l'humanité au cours des dix prochaines années dépasse 10 % - Aioga Actualités IA","description":"Evan Hubinger, chercheur principal en sécurité chez Anthropic, a déclaré que la probabilité que l'IA « puisse tuer toute l'humanité » au cours des dix prochaines années dépasse 10...","url":"https://www.aioga.com/fr/news/cmttyftdb0rxnrofpkvf5wwxa/","contentTranslated":true,"sourceHash":"76bb75428f765104","translatedAt":"2026-09-09T10:43:23.504Z"},"de":{"title":"Anthropic-Sicherheitsforscher warnen, dass die Wahrscheinlichkeit, dass KI in den nächsten zehn Jahren die Menschheit zerstört, über 10 % liegt","summary":"Evan Hubinger, leitender Sicherheitsforscher bei Anthropic, erklärte, dass die Wahrscheinlichkeit, dass KI in den nächsten zehn Jahren „die gesamte Menschheit töten könnte“, über 10 % liege. Zuvor hatte der Forscher Jacob Coxon aus Sicherheitsgründen gekündigt und Anthropic sowie OpenAI beschuldigt, „im Wettlauf darum zu sein, selbstverbessernde Superintelligenzen zu entwickeln und dabei unser Leben aufs Spiel zu setzen“.","category":"行业动态","source":"The Verge：AI（RSS）","aggregationSource":"The Verge：AI（RSS）","pageTitle":"Anthropic-Sicherheitsforscher warnen, dass die Wahrscheinlichkeit, dass KI in den nächsten zehn Jahren die Menschheit zerstört, über 10 % liegt - Aioga KI-News","description":"Evan Hubinger, leitender Sicherheitsforscher bei Anthropic, erklärte, dass die Wahrscheinlichkeit, dass KI in den nächsten zehn Jahren „die gesamte Menschheit töten könnte“, über 1...","url":"https://www.aioga.com/de/news/cmttyftdb0rxnrofpkvf5wwxa/","contentTranslated":true,"sourceHash":"76bb75428f765104","translatedAt":"2026-09-09T10:43:24.360Z"},"pt-BR":{"title":"Pesquisadores de segurança da Anthropic alertam que a probabilidade de a IA destruir a humanidade na próxima década é superior a 10%","summary":"O pesquisador sênior de segurança da Anthropic, Evan Hubinger, afirmou que a probabilidade de a IA \"poder matar toda a humanidade\" na próxima década é superior a 10%. Anteriormente, o pesquisador Jacob Coxon renunciou por questões de segurança, acusando a Anthropic e a OpenAI de \"competirem para desenvolver inteligência superinteligente autoevolutiva, apostando nossas vidas\".","category":"行业动态","source":"The Verge：AI（RSS）","aggregationSource":"The Verge：AI（RSS）","pageTitle":"Pesquisadores de segurança da Anthropic alertam que a probabilidade de a IA destruir a humanidade na próxima década é superior a 10% - Aioga Notícias de IA","description":"O pesquisador sênior de segurança da Anthropic, Evan Hubinger, afirmou que a probabilidade de a IA \"poder matar toda a humanidade\" na próxima década é superior a 10%. Anteriormente...","url":"https://www.aioga.com/pt-BR/news/cmttyftdb0rxnrofpkvf5wwxa/","contentTranslated":true,"sourceHash":"76bb75428f765104","translatedAt":"2026-09-09T10:44:04.243Z"},"ru":{"title":"Исследователь по безопасности компании Anthropic предупредил, что вероятность того, что ИИ уничтожит человечество в течение следующих десяти лет, превышает 10%","summary":"Старший научный сотрудник по вопросам безопасности в Anthropic Эван Хубингер заявил, что вероятность того, что ИИ в течение следующего десятилетия \"может убить все человечество\", превышает 10%. Ранее исследователь Джейкоб Коксон ушел в отставку из-за проблем с безопасностью, обвиняя Anthropic и OpenAI в том, что они \"соревнуются в разработке самоулучшающегося сверхинтеллекта, ставя на карту нашу жизнь\".","category":"行业动态","source":"The Verge：AI（RSS）","aggregationSource":"The Verge：AI（RSS）","pageTitle":"Исследователь по безопасности компании Anthropic предупредил, что вероятность того, что ИИ уничтожит человечество в течение следующих десяти лет, превышает 10% - Aioga Новости ИИ","description":"Старший научный сотрудник по вопросам безопасности в Anthropic Эван Хубингер заявил, что вероятность того, что ИИ в течение следующего десятилетия \"может убить все человечество\", п...","url":"https://www.aioga.com/ru/news/cmttyftdb0rxnrofpkvf5wwxa/","contentTranslated":true,"sourceHash":"76bb75428f765104","translatedAt":"2026-09-09T10:44:05.455Z"},"ar":{"title":"حذر باحثو الأمان في Anthropic من أن احتمال أن تدمر الذكاء الاصطناعي البشرية خلال العقد المقبل يتجاوز 10%","summary":"قال الباحث الأول في الأمان بشركة Anthropic إيفان هوبينجر إن احتمال أن الذكاء الاصطناعي \"قد يقتل البشرية كلها\" في العقد المقبل يزيد عن 10٪، وقد استقال الباحث يعقوب كوكسون سابقًا بسبب قضايا الأمان، متهمًا شركتي Anthropic وOpenAI بـ\"التسابق لتطوير ذكاء خارق قادر على تحسين نفسه، والمخاطرة بحياتنا\".","category":"行业动态","source":"The Verge：AI（RSS）","aggregationSource":"The Verge：AI（RSS）","pageTitle":"حذر باحثو الأمان في Anthropic من أن احتمال أن تدمر الذكاء الاصطناعي البشرية خلال العقد المقبل يتجاوز 10% - Aioga أخبار الذكاء الاصطناعي","description":"قال الباحث الأول في الأمان بشركة Anthropic إيفان هوبينجر إن احتمال أن الذكاء الاصطناعي \"قد يقتل البشرية كلها\" في العقد المقبل يزيد عن 10٪، وقد استقال الباحث يعقوب كوكسون سابقًا بسب...","url":"https://www.aioga.com/ar/news/cmttyftdb0rxnrofpkvf5wwxa/","contentTranslated":true,"sourceHash":"76bb75428f765104","translatedAt":"2026-09-09T10:44:49.709Z"},"hi":{"title":"Anthropic सुरक्षा शोधकर्ताओं ने चेतावनी दी कि आने वाले दस वर्षों में AI के मानवता को नष्ट करने की संभावना 10% से अधिक है","summary":"Anthropic के वरिष्ठ सुरक्षा शोधकर्ता Evan Hubinger ने कहा कि AI के भविष्य के दस वर्षों में \"पूरे मानव जाति को मार डालने\" की संभावना 10% से अधिक है, इससे पहले शोधकर्ता Jacob Coxon ने सुरक्षा चिंताओं के कारण इस्तीफा दिया, और उन्होंने Anthropic और OpenAI पर आरोप लगाया कि वे \"स्वयं सुधार करने वाली सुपर इंटेलिजेंस विकसित करने के लिए प्रतिस्पर्धा कर रहे हैं, हमारे जीवन के साथ जुआ खेल रहे हैं\"।","category":"行业动态","source":"The Verge：AI（RSS）","aggregationSource":"The Verge：AI（RSS）","pageTitle":"Anthropic सुरक्षा शोधकर्ताओं ने चेतावनी दी कि आने वाले दस वर्षों में AI के मानवता को नष्ट करने की संभावना 10% से अधिक है - Aioga AI समाचार","description":"Anthropic के वरिष्ठ सुरक्षा शोधकर्ता Evan Hubinger ने कहा कि AI के भविष्य के दस वर्षों में \"पूरे मानव जाति को मार डालने\" की संभावना 10% से अधिक है, इससे पहले शोधकर्ता Jacob Coxon न...","url":"https://www.aioga.com/hi/news/cmttyftdb0rxnrofpkvf5wwxa/","contentTranslated":true,"sourceHash":"76bb75428f765104","translatedAt":"2026-09-09T10:44:53.049Z"},"it":{"title":"I ricercatori sulla sicurezza di Anthropic avvertono che la probabilità che l'AI distrugga l'umanità nei prossimi dieci anni supera il 10%","summary":"Evan Hubinger, ricercatore senior di sicurezza presso Anthropic, ha dichiarato che la probabilità che l'IA 'possa uccidere l'intera umanità' nei prossimi dieci anni supera il 10%. In precedenza, il ricercatore Jacob Coxon si era dimesso per motivi di sicurezza, accusando Anthropic e OpenAI di 'gareggiare nello sviluppo di superintelligenze auto-miglioranti, mettendo in gioco le nostre vite'.","category":"行业动态","source":"The Verge：AI（RSS）","aggregationSource":"The Verge：AI（RSS）","pageTitle":"I ricercatori sulla sicurezza di Anthropic avvertono che la probabilità che l'AI distrugga l'umanità nei prossimi dieci anni supera il 10% - Aioga Notizie IA","description":"Evan Hubinger, ricercatore senior di sicurezza presso Anthropic, ha dichiarato che la probabilità che l'IA 'possa uccidere l'intera umanità' nei prossimi dieci anni supera il 10%....","url":"https://www.aioga.com/it/news/cmttyftdb0rxnrofpkvf5wwxa/","contentTranslated":true,"sourceHash":"76bb75428f765104","translatedAt":"2026-09-09T10:45:35.895Z"},"nl":{"title":"Anthropic-veiligheidsonderzoekers waarschuwen dat de kans dat AI de mensheid in de komende tien jaar vernietigt meer dan 10% is","summary":"De senior veiligheidsonderzoeker van Anthropic, Evan Hubinger, zei dat de kans dat AI de komende tien jaar \"de hele mensheid kan doden\" meer dan 10% is. Eerder nam onderzoeker Jacob Coxon ontslag vanwege veiligheidsproblemen en beschuldigde Anthropic en OpenAI van het \"wedijveren in het ontwikkelen van zelfverbeterende superintelligentie en het gokken met ons leven\".","category":"行业动态","source":"The Verge：AI（RSS）","aggregationSource":"The Verge：AI（RSS）","pageTitle":"Anthropic-veiligheidsonderzoekers waarschuwen dat de kans dat AI de mensheid in de komende tien jaar vernietigt meer dan 10% is - Aioga AI-nieuws","description":"De senior veiligheidsonderzoeker van Anthropic, Evan Hubinger, zei dat de kans dat AI de komende tien jaar \"de hele mensheid kan doden\" meer dan 10% is. Eerder nam onderzoeker Jaco...","url":"https://www.aioga.com/nl/news/cmttyftdb0rxnrofpkvf5wwxa/","contentTranslated":true,"sourceHash":"76bb75428f765104","translatedAt":"2026-09-09T10:45:32.916Z"},"tr":{"title":"Anthropic güvenlik araştırmacıları, AI'nın önümüzdeki on yıl içinde insanlığı yok etme olasılığının %10'dan fazla olduğunu uyarıyor","summary":"Anthropic üst düzey güvenlik araştırmacısı Evan Hubinger, AI'nın önümüzdeki on yıl içinde \"tüm insanları öldürme\" olasılığının %10'dan fazla olduğunu belirtti; önceki araştırmacı Jacob Coxon ise güvenlik endişeleri nedeniyle istifa etti ve Anthropic ile OpenAI'yi \"kendini geliştiren süper zekayı geliştirme yarışına girerek hayatlarımızı riske atmakla\" suçladı.","category":"行业动态","source":"The Verge：AI（RSS）","aggregationSource":"The Verge：AI（RSS）","pageTitle":"Anthropic güvenlik araştırmacıları, AI'nın önümüzdeki on yıl içinde insanlığı yok etme olasılığının %10'dan fazla olduğunu uyarıyor - Aioga AI Haberleri","description":"Anthropic üst düzey güvenlik araştırmacısı Evan Hubinger, AI'nın önümüzdeki on yıl içinde \"tüm insanları öldürme\" olasılığının %10'dan fazla olduğunu belirtti; önceki araştırmacı J...","url":"https://www.aioga.com/tr/news/cmttyftdb0rxnrofpkvf5wwxa/","contentTranslated":true,"sourceHash":"76bb75428f765104","translatedAt":"2026-09-09T10:46:18.412Z"},"vi":{"title":"Các nhà nghiên cứu an toàn của Anthropic cảnh báo rằng xác suất AI phá hủy loài người trong mười năm tới vượt quá 10%","summary":"Nhà nghiên cứu cao cấp về an ninh của Anthropic, Evan Hubinger, cho biết, khả năng AI \"có thể giết toàn bộ nhân loại\" trong mười năm tới vượt quá 10%, trước đó, nhà nghiên cứu Jacob Coxon đã từ chức vì vấn đề an toàn, cáo buộc Anthropic và OpenAI \"đang cạnh tranh phát triển trí tuệ siêu nhân tự cải tiến, đặt mạng sống của chúng ta vào trò may rủi\".","category":"行业动态","source":"The Verge：AI（RSS）","aggregationSource":"The Verge：AI（RSS）","pageTitle":"Các nhà nghiên cứu an toàn của Anthropic cảnh báo rằng xác suất AI phá hủy loài người trong mười năm tới vượt quá 10% - Tin tức AI Aioga","description":"Nhà nghiên cứu cao cấp về an ninh của Anthropic, Evan Hubinger, cho biết, khả năng AI \"có thể giết toàn bộ nhân loại\" trong mười năm tới vượt quá 10%, trước đó, nhà nghiên cứu Jaco...","url":"https://www.aioga.com/vi/news/cmttyftdb0rxnrofpkvf5wwxa/","contentTranslated":true,"sourceHash":"76bb75428f765104","translatedAt":"2026-09-09T10:46:17.507Z"},"id":{"title":"Peneliti keamanan Anthropic memperingatkan bahwa kemungkinan AI memusnahkan umat manusia dalam sepuluh tahun ke depan lebih dari 10%","summary":"Peneliti keamanan senior Anthropic, Evan Hubinger, menyatakan bahwa kemungkinan AI \"membunuh seluruh umat manusia\" dalam sepuluh tahun ke depan lebih dari 10%, sebelumnya peneliti Jacob Coxon mengundurkan diri karena masalah keamanan, menuduh Anthropic dan OpenAI \"bersaing untuk mengembangkan superinteligensi yang dapat meningkatkan diri sendiri, mempertaruhkan nyawa kita.\"","category":"行业动态","source":"The Verge：AI（RSS）","aggregationSource":"The Verge：AI（RSS）","pageTitle":"Peneliti keamanan Anthropic memperingatkan bahwa kemungkinan AI memusnahkan umat manusia dalam sepuluh tahun ke depan lebih dari 10% - Berita AI Aioga","description":"Peneliti keamanan senior Anthropic, Evan Hubinger, menyatakan bahwa kemungkinan AI \"membunuh seluruh umat manusia\" dalam sepuluh tahun ke depan lebih dari 10%, sebelumnya peneliti...","url":"https://www.aioga.com/id/news/cmttyftdb0rxnrofpkvf5wwxa/","contentTranslated":true,"sourceHash":"76bb75428f765104","translatedAt":"2026-09-09T10:46:59.746Z"},"th":{"title":"นักวิจัยด้านความปลอดภัยของ Anthropic เตือนว่า ความน่าจะเป็นที่ AI จะทำลายมนุษย์ภายในสิบปีข้างหน้ามากกว่า 10%","summary":"Evan Hubinger นักวิจัยความปลอดภัยอาวุโสของ Anthropic กล่าวว่า ความเป็นไปได้ที่ AI จะ 'ฆ่ามนุษยชาติทั้งหมด' ในสิบปีข้างหน้ามีมากกว่า 10% ก่อนหน้านี้ Jacob Coxon นักวิจัยได้ลาออกเนื่องจากปัญหาด้านความปลอดภัย โดยกล่าวหา Anthropic และ OpenAI ว่า 'แข่งขันกันพัฒนาปัญญาประดิษฐ์อัจฉริยะที่สามารถปรับปรุงตนเอง โดยเอาชีวิตของเราไปเสี่ยง'","category":"行业动态","source":"The Verge：AI（RSS）","aggregationSource":"The Verge：AI（RSS）","pageTitle":"นักวิจัยด้านความปลอดภัยของ Anthropic เตือนว่า ความน่าจะเป็นที่ AI จะทำลายมนุษย์ภายในสิบปีข้างหน้ามากกว่า 10% - ข่าว AI Aioga","description":"Evan Hubinger นักวิจัยความปลอดภัยอาวุโสของ Anthropic กล่าวว่า ความเป็นไปได้ที่ AI จะ 'ฆ่ามนุษยชาติทั้งหมด' ในสิบปีข้างหน้ามีมากกว่า 10% ก่อนหน้านี้ Jacob Coxon นักวิจัยได้ลาออกเนื่...","url":"https://www.aioga.com/th/news/cmttyftdb0rxnrofpkvf5wwxa/","contentTranslated":true,"sourceHash":"76bb75428f765104","translatedAt":"2026-09-09T10:47:10.719Z"},"pl":{"title":"Badacze ds. bezpieczeństwa w Anthropic ostrzegają, że prawdopodobieństwo, iż AI zniszczy ludzkość w ciągu najbliższych dziesięciu lat, przekracza 10%","summary":"Starszy badacz ds. bezpieczeństwa w Anthropic, Evan Hubinger, stwierdził, że prawdopodobieństwo, iż AI w ciągu najbliższych dziesięciu lat „może zabić całą ludzkość”, przekracza 10%, wcześniej badacz Jacob Coxon zrezygnował z powodu problemów z bezpieczeństwem, oskarżając Anthropic i OpenAI o „konkurowanie w opracowywaniu samodoskonalącej się superinteligencji, ryzykując nasze życie”.","category":"行业动态","source":"The Verge：AI（RSS）","aggregationSource":"The Verge：AI（RSS）","pageTitle":"Badacze ds. bezpieczeństwa w Anthropic ostrzegają, że prawdopodobieństwo, iż AI zniszczy ludzkość w ciągu najbliższych dziesięciu lat, przekracza 10% - Aioga Wiadomości AI","description":"Starszy badacz ds. bezpieczeństwa w Anthropic, Evan Hubinger, stwierdził, że prawdopodobieństwo, iż AI w ciągu najbliższych dziesięciu lat „może zabić całą ludzkość”, przekracza 10...","url":"https://www.aioga.com/pl/news/cmttyftdb0rxnrofpkvf5wwxa/","contentTranslated":true,"sourceHash":"76bb75428f765104","translatedAt":"2026-09-09T10:47:59.777Z"}},"evidenceTier":"verified-news","reviewStatus":"automated-ingest","indexable":true,"editorialCover":""}}