{ "version": 1, "event_id": "evt_b0831ae7fb72a1c8", "url": "https://xiyu.news/events/evt_b0831ae7fb72a1c8/", "json": "https://xiyu.news/api/events/evt_b0831ae7fb72a1c8.json", "type": "governance", "status": "monitoring", "category": "technology", "title": { "zh": "OpenAI解雇三名安全研究员并回应争议", "en": "OpenAI fires three safety researchers and responds to controversy" }, "current_state": { "zh": "OpenAI公开为其解雇安全研究员Jasmine Wang、Tomek Korbak和Mikita Balesni的决定辩护，称调查发现他们严重违背信任，并违反了处理敏感信息的明确政策。", "en": "OpenAI publicly stands by its firing of safety researchers Jasmine Wang, Tomek Korbak and Mikita Balesni, saying an investigation found they committed a significant breach of trust and violated clear policies on handling sensitive information." }, "first_seen_at": "2026-10-01T18:31:41.110947+00:00", "last_updated_at": "2026-10-09T11:21:04.012373+00:00", "last_material_change_at": "2026-10-09T11:21:04.012373+00:00", "confidence": 0.75, "updates_count": 6, "sources_count": 8, "entities": [ "ai-safety", "alleged", "fired", "fires", "group", "leak", "openai", "outside", "over", "researchers", "safety", "three", "wsj" ], "identifiers": [], "topics": [ "ai-governance", "ai-labs", "ai-policy", "ai-safety", "frontier-labs", "frontier-models", "information-security", "lab-accountability", "labor-and-personnel", "labor-dispute", "labor-turnover", "openai", "personnel", "personnel-changes", "third-party-evaluation" ], "updates": [ { "update_id": "upd_379a274b9af2097f", "event_id": "evt_b0831ae7fb72a1c8", "occurred_at": "2026-10-01T18:14:42Z", "published_at": "2026-10-01T18:14:42Z", "first_seen_at": "2026-10-01T18:31:41.110947Z", "time_precision": "published", "update_type": "initial", "material_change": true, "title_zh": "《华尔街日报》报道：OpenAI 与 3 名安全研究员断绝关系", "title_en": "OpenAI cuts ties with 3 safety researchers, WSJ reports", "what_changed_zh": "OpenAI 在一次内部调查后与三名安全研究员分道扬镳，调查结论认定他们不当处理了公司敏感信息。", "what_changed_en": "OpenAI separated from three safety researchers following an internal investigation that concluded they mishandled sensitive company information.", "current_state_zh": "OpenAI 在一次内部调查后与三名安全研究员分道扬镳，调查结论认定他们不当处理了公司敏感信息。", "current_state_en": "OpenAI separated from three safety researchers following an internal investigation that concluded they mishandled sensitive company information.", "detailed_summary_zh": "OpenAI separated from three safety researchers following an internal investigation that concluded they mishandled sensitive company information.", "detailed_summary_en": "OpenAI separated from three safety researchers following an internal investigation that concluded they mishandled sensitive company information.", "background_zh": "", "background_en": "", "community_discussion_zh": "", "community_discussion_en": "", "market_impact_zh": "", "market_impact_en": "", "importance_score": 7.5, "references": [], "confidence": 0.75, "story_ids": [ "rss:techcrunch.com_category_artificial-intelligence_feed_:08e08643693a8dbd", "rss:decrypt.co_feed:a471b58077e93529" ], "sources": [ { "url": "https://techcrunch.com/2026/10/01/openai-cuts-ties-with-three-safety-researchers-wsj-reports/", "label": "TechCrunch AI", "source_type": "rss", "official": false }, { "url": "https://decrypt.co/379857/openai-fires-researchers-alledged-leak-safety-group", "label": "Decrypt", "source_type": "rss", "official": false } ] }, { "update_id": "upd_f5d233f28e3ee2c6", "event_id": "evt_b0831ae7fb72a1c8", "occurred_at": "2026-10-03T04:41:24Z", "published_at": "2026-10-03T04:41:24Z", "first_seen_at": "2026-10-03T06:21:08.928970Z", "time_precision": "published", "update_type": "confirmation", "material_change": true, "title_zh": "OpenAI安全团队持续地震！负责人离职，三名员工因泄密被开", "title_en": "OpenAI安全团队持续地震！负责人离职，三名员工因泄密被开", "what_changed_zh": "OpenAI 安全团队持续动荡，安全部门负责人离职，三名员工因泄露信息被解雇。", "what_changed_en": "OpenAI's safety team is undergoing continued upheaval, with the head of the safety organization departing and three employees fired for leaking information.", "current_state_zh": "OpenAI 安全团队持续动荡，安全部门负责人离职，三名员工因泄露信息被解雇。", "current_state_en": "OpenAI's safety team is undergoing continued upheaval, with the head of the safety organization departing and three employees fired for leaking information.", "detailed_summary_zh": "OpenAI's safety team is undergoing continued upheaval, with the head of the safety organization departing and three employees fired for leaking information.", "detailed_summary_en": "OpenAI's safety team is undergoing continued upheaval, with the head of the safety organization departing and three employees fired for leaking information.", "background_zh": "", "background_en": "", "community_discussion_zh": "", "community_discussion_en": "", "market_impact_zh": "", "market_impact_en": "", "importance_score": 7.5, "references": [], "confidence": 0.75, "story_ids": [ "rss:www.qbitai.com_feed:4c2506b18ff43934" ], "sources": [ { "url": "https://www.qbitai.com/2026/10/501368.html", "label": "QbitAI 量子位", "source_type": "rss", "official": false } ] }, { "update_id": "upd_b09044f7fc03836f", "event_id": "evt_b0831ae7fb72a1c8", "occurred_at": "2026-10-03T14:31:56Z", "published_at": "2026-10-03T14:31:56Z", "first_seen_at": "2026-10-03T16:35:08.035729Z", "time_precision": "published", "update_type": "escalation", "material_change": true, "title_zh": "OpenAI一名安全部门员工已辞职并发出警告", "title_en": "An OpenAI safety employee has quit and is sounding the alarm", "what_changed_zh": "曾为OpenAI重大模型发布撰写安全报告的David Robinson本周辞职，并在《大西洋月刊》（The Atlantic）刊发评论文章，公开就AI安全风险发出警告。据有关其离职的报道，他在OpenAI工作约三年半，主导起草了现行的《前沿模型准备框架》（Preparedness Framework），并负责监督12次前沿模型发布的安全报告工作。\n\nRobinson是OpenAI重大模型发布配套安全文件的撰写者，因此他的公开批评会加剧外界对前沿实验室如何开展和披露安全监督的关注。有关其离职的报道称，他希望AI实验室借鉴航空与核能领域的安全实践。\n\n据报道，Robinson负责OpenAI的安全透明度工作，并参与了介绍其模型情况的系统卡（system card）的编写。有报道标题称他将OpenAI的文化形容为“已崩坏”，这一说法来自媒体的引述。", "what_changed_en": "David Robinson, who wrote the safety reports accompanying OpenAI's major model releases, has resigned and is publicly warning about AI safety risks in an editorial for The Atlantic.", "current_state_zh": "曾为OpenAI重大模型发布撰写安全报告的David Robinson本周辞职，并在《大西洋月刊》（The Atlantic）刊发评论文章，公开就AI安全风险发出警告。据有关其离职的报道，他在OpenAI工作约三年半，主导起草了现行的《前沿模型准备框架》（Preparedness Framework），并负责监督12次前沿模型发布的安全报告工作。\n\nRobinson是OpenAI重大模型发布配套安全文件的撰写者，因此他的公开批评会加剧外界对前沿实验室如何开展和披露安全监督的关注。有关其离职的报道称，他希望AI实验室借鉴航空与核能领域的安全实践。\n\n据报道，Robinson负责OpenAI的安全透明度工作，并参与了介绍其模型情况的系统卡（system card）的编写。有报道标题称他将OpenAI的文化形容为“已崩坏”，这一说法来自媒体的引述。", "current_state_en": "David Robinson, who wrote the safety reports accompanying OpenAI's major model releases, has resigned and is publicly warning about AI safety risks in an editorial for The Atlantic.", "detailed_summary_zh": "David Robinson, who wrote the safety reports accompanying OpenAI's major model releases, has resigned and is publicly warning about AI safety risks in an editorial for The Atlantic.", "detailed_summary_en": "David Robinson, who wrote the safety reports accompanying OpenAI's major model releases, has resigned and is publicly warning about AI safety risks in an editorial for The Atlantic.", "background_zh": "OpenAI发布《前沿模型准备框架》（Preparedness Framework），并在重大模型发布时配套发布记录风险评估与缓解措施的安全报告和系统卡（system card）。在辞职之前，Robinson的职责正围绕这一信息披露流程展开。", "background_en": "OpenAI publishes a Preparedness Framework and accompanies major model releases with safety reports and system cards that document evaluated risks and mitigations. Robinson's role centered on that disclosure pipeline before his resignation.", "community_discussion_zh": "评论意见分歧明显：有人质疑Robinson的动机，认为离职时机与股权归属期重合，且他聘请了公关公司；也有人认为业界应更关注当下具体问题，如沙箱隔离和模型有害输出，而非假想的未来风险。部分评论表示，在据称员工承受巨大压力的情况下，很难想象在OpenAI工作会愉快；还有人转而抱怨ChatGPT的界面与会话一致性问题，与本次事件无直接关系。", "community_discussion_en": "Commenters were divided: several questioned Robinson's motives, suggesting the timing coincided with vested equity and that he had retained a PR firm, while others argued the field needs more focus on present-day, concrete problems such as sandboxing and harmful model outputs rather than hypothetical future risks. Some said they doubted working at OpenAI is pleasant given the pressure employees are reported to face, and others shifted to unrelated complaints about ChatGPT's UI and conversation consistency.", "market_impact_zh": "这是一则人事与治理层面的消息，对加密市场没有直接的托管、流动性或供给传导渠道。若产生影响，也只会通过AI叙事类代币的市场情绪传导，而这类代币对有关前沿实验室如何管理安全监督的消息较为敏感。", "market_impact_en": "This is a personnel and governance story with no direct custody, liquidity, or supply channel into crypto markets. Any transmission would run only through sentiment around AI-narrative tokens, which can be sensitive to headlines about how frontier labs manage safety oversight.", "importance_score": 7.0, "references": [ { "url": "https://cellcog.ai/blog/openai-safety-lead-resigns/", "title": "OpenAI Safety Lead David Robinson Quits: His Essay, Read | CellCog" }, { "url": "https://sputnikglobe.com/20261003/openai-safety-head-david-robinson-resigns---reports-1124829799.html", "title": "OpenAI Safety Head David Robinson Resigns - Reports" }, { "url": "https://digg.com/tech/tzgk6ewq", "title": "OpenAI safety leader quits, calling its culture “broken” · Digg" } ], "confidence": 0.75, "story_ids": [ "rss:www.theverge.com_rss_ai-artificial-intelligence_index.xml:d9a8d4e28cd0f673" ], "sources": [ { "url": "https://www.theverge.com/ai-artificial-intelligence/1004408/openai-safety-quits-sounding-the-alarm", "label": "The Verge AI", "source_type": "rss", "official": false } ] }, { "update_id": "upd_261e3c9490644a53", "event_id": "evt_b0831ae7fb72a1c8", "occurred_at": "2026-10-03T22:18:13Z", "published_at": "2026-10-03T22:18:13Z", "first_seen_at": "2026-10-04T02:50:27.636990Z", "time_precision": "published", "update_type": "aftermath", "material_change": true, "title_zh": "OpenAI 安全负责人辞职，警告这家 AI 公司文化“已崩坏”", "title_en": "OpenAI safety leader quits, warning AI company's culture is 'broken'", "what_changed_zh": "OpenAI 一位高级安全负责人辞职，并公开警告公司内部文化“已崩坏”，引发关于 AI 安全工作是聚焦当下危害还是假设性未来风险的争论。", "what_changed_en": "A senior OpenAI safety leader resigned while publicly warning that the company's internal culture is 'broken', prompting debate about whether AI safety efforts are focused on present harms or hypothetical future risks.", "current_state_zh": "OpenAI 一位高级安全负责人辞职，并公开警告公司内部文化“已崩坏”，引发关于 AI 安全工作是聚焦当下危害还是假设性未来风险的争论。", "current_state_en": "A senior OpenAI safety leader resigned while publicly warning that the company's internal culture is 'broken', prompting debate about whether AI safety efforts are focused on present harms or hypothetical future risks.", "detailed_summary_zh": "A senior OpenAI safety leader resigned while publicly warning that the company's internal culture is 'broken', prompting debate about whether AI safety efforts are focused on present harms or hypothetical future risks.", "detailed_summary_en": "A senior OpenAI safety leader resigned while publicly warning that the company's internal culture is 'broken', prompting debate about whether AI safety efforts are focused on present harms or hypothetical future risks.", "background_zh": "", "background_en": "", "community_discussion_zh": "", "community_discussion_en": "", "market_impact_zh": "", "market_impact_en": "", "importance_score": 7.5, "references": [ { "url": "https://news.ycombinator.com/item?id=49948332", "title": "Community discussion" } ], "confidence": 0.75, "story_ids": [ "hackernews:story:49948332" ], "sources": [ { "url": "https://www.theguardian.com/technology/2026/oct/03/openai-safety-leader-quits-warning-ai-companys-culture-is-broken", "label": "jethronethro", "source_type": "hackernews", "official": false } ] }, { "update_id": "upd_3e24a3b5585b3019", "event_id": "evt_b0831ae7fb72a1c8", "occurred_at": "2026-10-08T20:04:26Z", "published_at": "2026-10-08T20:04:26Z", "first_seen_at": "2026-10-08T23:27:54.895638Z", "time_precision": "published", "update_type": "response", "material_change": true, "title_zh": "被解雇的OpenAI安全研究员反驳不当行为指控，并警告寒蝉效应", "title_en": "Fired OpenAI safety researchers dispute misconduct claims, warn of chilling effect", "what_changed_zh": "三名被解雇的OpenAI安全研究员公开否认有关其不当处理敏感信息的指控，并发表公开信警告解雇事件正在公司内部对AI安全工作产生寒蝉效应。", "what_changed_en": "Three fired OpenAI safety researchers publicly rejected the misconduct allegations and published an open letter warning that their dismissals are creating a chilling effect on AI safety work inside the company.", "current_state_zh": "OpenAI一位高级安全负责人辞职并警告公司文化“已崩坏”，三名被解雇的安全研究员则公开反驳不当行为指控，警告解雇事件正对AI安全工作产生寒蝉效应。", "current_state_en": "A senior OpenAI safety leader resigned warning the internal culture is 'broken', while the three fired safety researchers publicly disputed misconduct allegations and warned their dismissals are chilling AI safety work.", "detailed_summary_zh": "三名被OpenAI解雇的安全研究员公开否认有关其不当处理敏感信息的指控，并发表公开信反驳这些不当行为说法，同时警告他们的解雇正在公司内部对AI安全工作产生寒蝉效应。他们在信中写道：“鉴于AI发展所带来的重大安全担忧，员工不应在一个由恐惧和模糊规则阻碍AI安全工作、削弱第三方问责的环境中工作。”\n\n这封公开信将这些解雇事件界定为治理问题，而非单纯的人事问题，认为处理方式会影响这家领先AI实验室内部安全工作的开展以及第三方问责机制。\n\n这些研究员否认“不当共享敏感信息”的定性；此前的报道称，这些解雇与涉嫌向外部安全组织分享信息有关。", "detailed_summary_en": "Three fired OpenAI safety researchers publicly rejected allegations that they mishandled sensitive information, publishing an open letter that disputes the misconduct claims and warns their dismissals are chilling AI safety work inside the company. In the letter they wrote that \"given the significant safety concerns surrounding the development of AI, employees must not be left working in an environment where fear and unclear rules stymie AI safety work and weaken third-party accountability.\"\n\nThe letter frames the dismissals as a governance issue rather than a purely personnel matter, arguing that the way the cases were handled affects how safety work and third-party accountability function at a leading AI lab.\n\nThe researchers dispute the characterization that they improperly shared sensitive information; earlier reporting linked the dismissals to allegations of sharing information with external safety groups.", "background_zh": "OpenAI是一家总部位于旧金山的美国AI公益公司，开发GPT系列大语言模型。该公司此前已因安全团队配置而受到公众审视，有报道称其“使命对齐”（Mission Alignment）团队在成立约16个月后被解散。另外，据维基百科对2026年一起事件的记述，超过1100名前沿AI公司员工签署公开信，敦促美国政府建立有意放缓AI发展速度的机制。", "background_en": "OpenAI is an American AI public benefit corporation headquartered in San Francisco that develops the GPT series of large language models. The company has faced prior public scrutiny over its safety staffing, including reporting that its Mission Alignment team was disbanded roughly 16 months after being formed. Separately, more than 1,100 employees of frontier AI companies signed an open letter urging the US government to develop means of deliberately pacing AI development, according to a Wikipedia summary of a 2026 incident.", "community_discussion_zh": "", "community_discussion_en": "", "market_impact_zh": "该事件没有直接关联的资产敞口，但会强化围绕前沿AI实验室的治理与监管审视叙事；这一主题可能通过合规与声誉风险的认知，影响市场对AI相关代币以及AI开发企业股权敞口的情绪。", "market_impact_en": "The story has no direct asset exposure, but it feeds into the governance-and-regulatory-scrutiny narrative around frontier AI labs, a theme that can influence sentiment toward AI-related tokens and equity exposure to AI developers through perceived compliance and reputational risk.", "importance_score": 7.0, "references": [ { "url": "https://techcrunch.com/2026/10/08/fired-openai-safety-researchers-dispute-misconduct-claims-warn-of-chilling-effect/", "title": "Fired OpenAI safety researchers dispute misconduct... | TechCrunch" }, { "url": "https://www.techtimes.com/articles/328427/20261002/openai-fires-3-safety-researchers-allegedly-sharing-info-external-safety-groups-says-report.htm", "title": "OpenAI Fires 3 Safety Researchers For Allegedly Sharing Info to..." } ], "confidence": 0.95, "story_ids": [ "rss:techcrunch.com_category_artificial-intelligence_feed_:c7bf63318b0ab261" ], "sources": [ { "url": "https://techcrunch.com/2026/10/08/fired-openai-safety-researchers-dispute-misconduct-claims-warn-of-chilling-effect/", "label": "TechCrunch AI", "source_type": "rss", "official": false } ] }, { "update_id": "upd_69344c3fe552e0f4", "event_id": "evt_b0831ae7fb72a1c8", "occurred_at": "2026-10-09T06:44:02Z", "published_at": "2026-10-09T06:44:02Z", "first_seen_at": "2026-10-09T07:11:46.878622Z", "time_precision": "published", "update_type": "confirmation", "material_change": true, "title_zh": "OpenAI回应安全研究人员离职：因敏感信息违规，并非针对提出安全担忧行为", "title_en": "OpenAI回应安全研究人员离职：因敏感信息违规，并非针对提出安全担忧行为", "what_changed_zh": "OpenAI表示，其解雇三名前员工Jasmine、Mikita和Tomek，是因为他们违反敏感信息处理政策，并存在超出其公开信所披露范围的严重背信行为；OpenAI否认此举是对他们提出安全担忧的报复，同时表示正在敲定与第三方安全评估机构的合同，并将在未来几周内公布。", "what_changed_en": "OpenAI said it terminated three former employees — Jasmine, Mikita and Tomek — for violating sensitive-information handling policies and serious breaches of trust beyond what their public letter disclosed, denying the move was retaliation for raising safety concerns, while stating it is finalizing contracts with third-party safety evaluators to be announced in the coming weeks.", "current_state_zh": "OpenAI表示，其解雇三名前员工Jasmine、Mikita和Tomek，是因为他们违反敏感信息处理政策，并存在超出其公开信所披露范围的严重背信行为；OpenAI否认此举是对他们提出安全担忧的报复，同时表示正在敲定与第三方安全评估机构的合同，并将在未来几周内公布。", "current_state_en": "OpenAI said it terminated three former employees — Jasmine, Mikita and Tomek — for violating sensitive-information handling policies and serious breaches of trust beyond what their public letter disclosed, denying the move was retaliation for raising safety concerns, while stating it is finalizing contracts with third-party safety evaluators to be announced in the coming weeks.", "detailed_summary_zh": "OpenAI said it terminated three former employees — Jasmine, Mikita and Tomek — for violating sensitive-information handling policies and serious breaches of trust beyond what their public letter disclosed, denying the move was retaliation for raising safety concerns, while stating it is finalizing contracts with third-party safety evaluators to be announced in the coming weeks.", "detailed_summary_en": "OpenAI said it terminated three former employees — Jasmine, Mikita and Tomek — for violating sensitive-information handling policies and serious breaches of trust beyond what their public letter disclosed, denying the move was retaliation for raising safety concerns, while stating it is finalizing contracts with third-party safety evaluators to be announced in the coming weeks.", "background_zh": "三人是此前发布公开信的前 OpenAI 安全研究人员，公开信警告他们的解雇可能对 AI 安全工作产生寒蝉效应；OpenAI 对 CNN 表示，解雇源于违反敏感信息政策的特定行为，而非安全倡议。Korbak 曾发表关于思维链可监控性的研究，而 OpenAI 称这一方向是其长期研究重点。第三方安全评估在前沿实验室中仍属新兴做法——据检索结果，Anthropic 曾指定埃森哲为其首个嵌入式评估方。", "background_en": "The three researchers are among the former OpenAI safety staff who published an open letter warning that their firings could have a chilling effect on AI safety work; OpenAI told CNN the dismissals stemmed from specific conduct violating its sensitive-information policies rather than safety advocacy. Korbak has published research on chain-of-thought monitorability, a theme OpenAI cites as a long-running research focus. Third-party safety evaluation is an emerging practice at frontier labs — per search results, Anthropic named Accenture as its first embedded evaluator.", "community_discussion_zh": "", "community_discussion_en": "", "market_impact_zh": "这一人事争议对加密资产、交易场所或托管安排没有直接敞口；可能的影响仅通过市场情绪间接传导至跟随大型 AI 实验室新闻波动的 AI 叙事类代币。声明中的第三方评估承诺属于治理层面事项，尚无明确对应链上或代币层面的传导机制。", "market_impact_en": "There is no direct exposure of crypto assets, venues or custody arrangements to this personnel dispute; any effect would run indirectly through sentiment around AI-narrative tokens that trade on news flow from major AI labs. The statement's third-party evaluation commitment is a governance item with no defined on-chain or token-level mechanism.", "importance_score": 7.5, "references": [ { "url": "https://edition.cnn.com/2026/10/08/tech/fired-open-ai-researchers-pushed-out", "title": "Fired OpenAI safety researchers say they were pushed out over..." }, { "url": "https://www.gadgetreview.com/former-openai-researchers-warn-their-firings-could-have-a-chilling-effect-on-ai-safety", "title": "Former OpenAI Researchers Warn Their Firings... - Gadget Review" }, { "url": "https://tomekkorbak.com/cot-monitorability-is-a-fragile-opportunity/cot_monitoring.pdf", "title": "Chain of Thought Monitorability" } ], "confidence": 0.75, "story_ids": [ "telegram:theblockbeats:199644", "rss:www.theverge.com_rss_ai-artificial-intelligence_index.xml:ec97dad83a3aa87e" ], "sources": [ { "url": "https://m.theblockbeats.info/flash/370964?from=telegram", "label": "theblockbeats", "source_type": "telegram", "official": false }, { "url": "https://www.theverge.com/ai-artificial-intelligence/1008604/openai-defends-decision-fire-safety-researchers", "label": "The Verge AI", "source_type": "rss", "official": false } ] } ] }