智通财经获悉，美国总统特朗普下令联邦政府机构停止使用人工智能公司Anthropic的产品，使这家AI 企业与五角大楼之间围绕技术使用限制的争端进一步升级。 特朗普周五在社交媒体上表示，已指示"所有联邦机构立即停止使用Anthropic技术"，并称若该公司不配 合过渡安排，可能面临"重大的民事和刑事后果"。他同时宣布，包括国防部在内的相关机构将有六个月 的"过渡期"逐步停用Anthropic产品。 特朗普的决定也可能在硅谷引发反弹。此前包括亚马逊(AMZN.US)和微软(MSFT.US)在内的多家科技公 司员工曾呼吁管理层拒绝五角大楼提出的"无限制使用"要求。Anthropic方面周四重申立场称："这些威 胁不会改变我们的决定，我们无法违背良知接受相关要求。" 财经频道更多独家策划、专家专栏，免费查阅>> 责任编辑：山上 特朗普此举预计将对AI行业产生震动。Anthropic此前与军方签署的合同规模最高可达2亿美元，此外还 为国务院等民事机构提供服务。若全面终止合作，将削弱公司在联邦市场的布局。当前Anthropic估值 高达约3800亿美元，外界普遍预期其最快可能于今年启动首次公开募股(IPO)。 值得注意的 ...

美国总统特朗普下令联邦政府机构停止使用人工智能公司Anthropic的产品，使这家AI企业与五角大楼 之间围绕技术使用限制的争端进一步升级。 特朗普周五在社交媒体上表示，已指示"所有联邦机构立即停止使用Anthropic技术"，并称若该公司不配 合过渡安排，可能面临"重大的民事和刑事后果"。他同时宣布，包括国防部在内的相关机构将有六个月 的"过渡期"逐步停用Anthropic产品。 值得注意的是，Anthropic此前是少数能够在五角大楼机密云环境中运行的AI系统之一，其"Claude Gov"工具在部分国防人员中使用较广。若被完全排除在政府体系之外，可能在短期内带来一定的国家 安全和技术替代挑战。 与此同时，竞争对手正加速争夺政府业务。马斯克旗下xAI已获准参与涉密项目，OpenAI及谷歌 (GOOG.US,GOOGL.US)的Gemini模型也在积极拓展联邦市场。OpenAI首席执行官奥特曼在内部备忘录 中表示，公司正与国防部沟通，在设定一定限制的前提下提供模型服务。 特朗普的决定也可能在硅谷引发反弹。此前包括亚马逊(AMZN.US)和微软(MSFT.US)在内的多家科技公 司员工曾呼吁管理层拒绝五角大楼 ...

Core Insights - Responsible AI is transitioning from a practice of a few companies to an industry standard, with deeper collaboration emerging in the US-China AI ecosystem driven by the strategic foresight and practical actions of CIOs [1][9]. Group 1: CIO Role Evolution - The role of Chief Information Officers (CIOs) has evolved from traditional technology managers to core drivers of enterprise strategy, guardians of risk control, and bridges for cross-border technology cooperation [3]. - CIOs are now expected to balance innovation with risk management, requiring a shift in mindset to become strategic business enablers rather than just technical supporters [8][9]. Group 2: Responsible AI Adoption - Only 28% of US respondents view "responsible AI" as a core business priority, and only 33% have implemented clear applications across their organizations, indicating a significant gap in AI governance maturity [3][4]. - The rapid pace of AI technology evolution has outstripped the development of governance frameworks, leading to a low maturity level in responsible AI practices [4][10]. Group 3: Data Governance Importance - Data is recognized as the fuel for AI, with high-quality data being essential for generating valuable AI outcomes. Effective data governance is crucial for the successful implementation of responsible AI [7]. - Companies with established data governance frameworks see a 2.8 times higher success rate in AI projects compared to those without such frameworks [7]. Group 4: Global AI Regulation Perspectives - There are significant regional differences in AI regulation, with the US and China adopting a more relaxed approach compared to Europe and the Middle East, which favor stricter regulations [5]. - The EU's AI Act introduces stringent compliance requirements for high-risk AI systems, which can inhibit innovation, particularly for small and medium-sized enterprises [5]. Group 5: Multi-AI Model Strategy - A multi-AI model strategy is essential for global enterprises to navigate varying regulatory requirements and business needs across different regions [9]. - Companies must adapt their AI model choices based on local compliance and operational demands, ensuring flexibility in their AI deployments [9]. Group 6: Future of AI in Business - The future of AI is seen as a dual opportunity and challenge for CIOs, who must navigate technological advancements, regulatory differences, and data governance to drive responsible AI development [9]. - As AI technology continues to evolve, responsible AI practices are expected to become standard across industries, fostering deeper collaboration in the US-China AI ecosystem [9].

Group 1 - The role of Chief Information Officers (CIOs) has evolved from traditional technology managers to core drivers of enterprise strategy, guardians of risk control, and bridges for cross-border technology collaboration [2][3] - A recent PwC survey indicates that only 28% of U.S. respondents view "responsible AI" as a top business priority, and only 33% of companies have implemented clear applications across the organization [2][11] - McKinsey's 2024 global AI survey shows that while about 60% of companies have initiated AI projects, only 15% have established comprehensive AI governance frameworks, with average returns on AI investments falling short of the expected 30% [2][3] Group 2 - Responsible AI should not only focus on risk mitigation but also on helping businesses extract more commercial value from AI systems, transforming it from a compliance tool to a value extraction engine [3][4] - The low maturity of responsible AI practices is attributed to the imbalance between the rapid pace of technological iteration and the development of governance frameworks [3][4] - The emergence of AI agents has highlighted the inadequacy of traditional application management models, complicating the establishment of forward-looking governance frameworks [3][4] Group 3 - Global differences in AI regulation were discussed, with the U.S. and China seen as more relaxed compared to Europe and the Middle East, which adopt stricter regulatory approaches [4][5] - The EU AI Act categorizes AI systems by risk levels, imposing stringent compliance requirements on high-risk AI systems, which can inhibit innovation, particularly for small and medium enterprises [5][6] - A unified global AI standard is desired to reduce cross-border operational costs, similar to telecommunications standards [5][6] Group 4 - Data governance is crucial for responsible AI implementation, with high-quality data being essential for generating quality AI outcomes [6][7] - Companies must invest significant effort in data governance, ensuring proper data management and access control to prevent sensitive information leaks [6][7] - Organizations with established data governance frameworks see a 2.8 times higher success rate in AI projects compared to those without such frameworks [6][7] Group 5 - The evolution of the CIO role requires a balance of entrepreneurial spirit and a strong sense of responsibility, as they must drive innovation while safeguarding data security and compliance [7][8] - CIOs are now seen as strategic business enablers, leveraging core data assets to enhance productivity and differentiate business offerings [7][8] - The challenges posed by geopolitical uncertainties have led to a focus on "supply chain resilience" among global enterprises [7][8] Group 6 - The importance of a multi-AI model strategy was emphasized, as different AI models have varying service terms and usage restrictions, necessitating compliance with regional regulations [8][9] - CIOs must navigate the complexities of cross-border regulations while ensuring the selection of the most suitable AI models for their business needs [8][9] - The dual-supplier strategy is being adopted to mitigate risks associated with reliance on a single technology source [8][9] Group 7 - The rapid evolution of AI technology presents both opportunities and challenges for CIOs, who must adapt to changing landscapes and governance requirements [9][10] - The future of responsible AI is expected to shift from being a practice of a few companies to becoming an industry standard, driven by the strategic foresight and pragmatic actions of CIOs [9][10]

Core Viewpoint - The Hong Kong Stock Exchange (HKEX) emphasizes the importance of balancing innovation and stability in the context of AI, advocating for responsible AI practices to mitigate governance, privacy, and security risks [1][3]. Group 1: Responsible AI Best Practices - The five best practices for responsible AI suggested by HKEX include: 1. Establishing a robust AI governance framework [3] 2. Promoting human oversight to maintain a "human-machine collaboration" model [3] 3. Ensuring data quality and diversity to reduce bias [3] 4. Enhancing AI literacy and risk awareness across the organization [3] 5. Deploying AI securely through local or trusted cloud infrastructure [3] Group 2: Opportunities in the Greater Bay Area - The Greater Bay Area is recognized as a leading global technology innovation cluster, providing unique advantages for the development of technology in exchanges [3] - HKEX looks forward to deepening collaboration with peers in Shenzhen and Guangzhou to fully leverage AI's potential in capital markets and broader economic contexts, while maintaining market integrity and stability [3]