【导读】苹果公司被意大利监管机构罚超9800万欧元 中国基金报记者 忆山 苹果公司再因涉嫌滥用市场支配地位被重罚！ 12月22日，意大利竞争与市场管理局（AGCM）发布的消息显示，苹果公司因滥用市场支配 地位，被处以9863.5万欧元的罚款，约合8.1亿元人民币。 对此，苹果公司表示，强烈反对意大 利反垄断监管机构的决定，并将提起上诉。 应用跟踪透明度政策存在限制竞争行为 意大利竞争与市场管理局表示，经调查认定，苹果公司的应用跟踪透明度政策（ATT）存在限 制竞争行为。 意大利竞争与市场管理局指出，自2021年4月起，苹果公司针对iOS设备推行适用于应用商店 第三方应用开发者的应用跟踪透明度政策（ATT），从竞争法层面判定具有限制性竞争本质。 具体来看，第三方应用开发者若要通过苹果ATT弹窗，出于广告推广目的收集并关联用户数 据，必须事先获得用户的专项授权。然而，该弹窗的设计不符合隐私法规要求，导致开发者 不得不就同一数据收集目的，向用户重复发起授权申请。 该局认定，应用跟踪透明度政策（ATT）的条款由苹果公司单方面强制推行，损害了其商业合 作伙伴的合法权益。同时，该政策条款的严苛程度，与苹果声称的数据保护目 ...

Core Insights - The Doubao mobile assistant team announced adjustments to AI capabilities in response to public concerns and operational challenges following the launch of their new features [1][2][3] - The Doubao mobile assistant, developed in collaboration with ZTE, allows users to automate tasks across multiple applications, but has faced restrictions from various financial and gaming apps due to security and privacy concerns [1][2] Group 1: AI Capabilities and Adjustments - Doubao plans to implement regulatory adjustments to AI operations on mobile devices to ensure technology development aligns with industry acceptance and user experience [3] - The AI assistant's ability to perform tasks such as price comparison and order placement across platforms has garnered significant market attention, but has also raised issues regarding data privacy and security [2][3] Group 2: Restrictions Imposed by Other Apps - Major apps, including WeChat, Taobao, Alipay, and several banking applications, have begun to restrict or disable access to Doubao due to concerns over security and operational integrity [2] - The Doubao team acknowledges the need to limit AI capabilities in specific scenarios, particularly in financial applications, to safeguard user funds and comply with industry standards [4] Group 3: Specific Limitations on AI Usage - The company will restrict the use of AI in scenarios involving score manipulation and incentivized interactions to maintain the integrity of user engagement [4] - Financial applications will see a temporary suspension of AI operations to prioritize user security, despite the assistant requiring user authorization for sensitive actions [4] - Certain gaming scenarios will also have AI capabilities limited to ensure fairness in competitive rankings [4]

Core Viewpoint - The Texas Attorney General Ken Paxton is investigating online retailer Shein for potentially selling unsafe products and violating state laws regarding product safety and ethical sourcing [1][3]. Group 1: Investigation Details - The investigation aims to determine if Shein's supply chain and production practices violate Texas laws, including the use of toxic or harmful substances [1][3]. - Paxton emphasized that companies cutting corners on product safety will be held accountable, asserting that consumers have the right to know if the products they purchase are ethical, safe, and transparent [1][3]. Group 2: Consumer Protection - The investigation will also examine Shein's data collection and privacy protection measures, highlighting concerns over consumer rights and safety [1][3]. - Paxton stated that he will not allow cheap and dangerous foreign goods to enter the U.S., which could jeopardize public health [1][3].

Core Insights - The book "Data Capital" argues that data is a new form of public resource and a key production factor that can drive economic development, necessitating a reevaluation of data ownership and usage rights to achieve economic transformation [4][5]. Group 1: Data as a Production Factor - Data is recognized as a core production factor alongside land, labor, and financial capital, but it is predominantly controlled by a few entities, leading to data monopolies [6]. - The monopolization of data hinders economic growth and social progress, prompting a need to rethink data ownership and usage rights [6][7]. Group 2: Community Data Cooperation - The authors advocate for community data cooperation as a solution to data monopolization, where individuals voluntarily share their data to create value for community members [7]. - An example is provided of a data cooperative for artists and musicians, initiated by institutions like Berklee College of Music and MIT, aimed at improving the music ecosystem through technology and incentive mechanisms [7][8]. Group 3: Open Access and Metadata Models - The music industry can adopt open access paradigms from other sectors to create optional creative metadata models, which help reduce transaction complexity and operational costs [8][9]. - A three-layer architecture for the digital music ecosystem is proposed, including a music metadata layer, a copyright and royalty management layer, and a virtual asset layer for music rights [9]. Group 4: Privacy and Open Algorithms - The book introduces a five-layer architecture for data privacy protection, emphasizing the importance of designing shared analytical results rather than exporting data [10][11]. - Two privacy-preserving computation paradigms are discussed: secret sharing and multi-party computation, which enhance data security while allowing collaborative analysis [11]. Group 5: Blockchain and Interoperability - The development of blockchain technology faces challenges in achieving interoperability among different blockchain networks, which is essential for its role as a foundational infrastructure in future commerce [12][13]. - The authors emphasize that human considerations must be central to discussions about digital capital, aiming for a system that promotes health, inclusivity, and benefits for all stakeholders [13].

Core Viewpoint - The article discusses the emergence of Jinghua Misan, a company focused on privacy-preserving computing solutions, particularly in the context of AI and data security, highlighting its recent recognition and the potential market for its technology [3][5][22]. Group 1: Company Overview - Jinghua Misan, incubated at Tsinghua University, aims to address privacy concerns associated with cloud-based AI models by utilizing a secure computing engine that operates at acceptable costs [3][5]. - The company won the "DEMO GOD" award at the 2025 DEMO CHINA event, standing out among 126 tech companies [3][5]. Group 2: Technology and Solutions - Homomorphic encryption allows computations on encrypted data without decryption, maintaining data privacy throughout its lifecycle [7][9]. - Jinghua Misan's technology aims to reduce the performance loss typically associated with homomorphic encryption, making it viable for AI applications [9][12]. - The company is developing two main products: a homomorphic inference engine for AI models and a homomorphic training engine to protect data ownership during transactions [15][17]. Group 3: Market Demand and Challenges - There is a strong market demand for high-performance, practical homomorphic computing solutions, particularly in privacy-sensitive applications [8][11]. - Current solutions face challenges such as high computational costs and hardware dependencies, which limit their widespread adoption [9][12]. Group 4: Future Vision and Strategy - Jinghua Misan aims to focus on AI-related privacy solutions, believing that its technology can become a new consensus for secure data transactions and AI model usage [22]. - The company plans to launch its homomorphic inference and training engines by the end of this year and early next year, respectively [18].

Core Insights - The article emphasizes the importance of multi-channel interaction between businesses and customers in the digital marketing era, highlighting SCRM (Social Customer Relationship Management) systems as a key tool for automating and intelligentizing customer interactions [1] Group 1: Technical Architecture of SCRM Systems - The data collection layer of SCRM systems integrates all customer touchpoints, such as WeChat, Weibo, Douyin, and offline stores, enabling real-time capture of customer behavior data [3] - The data processing layer involves cleaning, analyzing, and tagging raw data, utilizing NLP technology to assess customer sentiment and RFM models to create comprehensive customer profiles [4] - The business application layer allows for automated marketing, intelligent customer service, and sales forecasting based on customer profiles [5][6][7] Group 2: Customer Interaction Process Design - The customer interaction process consists of four steps: channel integration and data cleaning, automated rule setting, personalized content pushing, and performance evaluation [9][10] - A hybrid service model is recommended to balance automation and human interaction, ensuring customer experience is not compromised [11] Group 3: Common Issues and Solutions in SCRM Implementation - Data silos can be addressed by selecting SCRM systems that support API integration and establishing a unified customer ID system [13] - Private traffic operations should avoid low engagement by segmenting customers based on RFM models and designing a comprehensive engagement strategy [13] - Compliance risks related to data privacy can be mitigated by implementing explicit consent features and regular compliance checks [13] Group 4: Optimization Strategies for SCRM Systems - Upgrading the technical architecture to support high-volume data queries is essential for efficiency [17] - Establishing cross-departmental collaboration mechanisms can enhance SCRM system adoption and effectiveness [17] - Continuous iteration based on feedback is crucial for optimizing system performance and user experience [17] Group 5: Industry Practices of SCRM Systems - In the retail sector, personalized recommendations through SCRM have led to a 20% increase in sales [20] - In the banking industry, customer segmentation using SCRM has improved customer satisfaction by 30% and increased sales of financial products by 18% [21] - B2B companies have accelerated deal closures by utilizing SCRM for customer feedback and support, resulting in a 25% increase in customer retention [21] Group 6: Future Trends of SCRM Systems - The future of SCRM systems is expected to be driven by advancements in AI and big data, focusing on smarter and more personalized customer interactions [23] - Companies that effectively utilize SCRM can achieve a 47% higher customer retention rate compared to their peers, emphasizing the need to break down data silos and foster cross-departmental collaboration [23]

Core Points - The EU General Court upheld the data transfer agreement between the EU and the US, rejecting a lawsuit by French MEP Philippe Latombe, who argued that the agreement did not fully respect EU data protection rules [1] - The court's ruling indicates that the US has ensured adequate protection for personal data transferred from the EU, which has been welcomed by US digital lobbying groups [1] - Privacy advocates expressed surprise at the court's decision, as they expected a procedural dismissal rather than a substantive validation of the agreement [1] Group 1: Legal Framework and Challenges - Since 1995, EU law prohibits the transfer of personal data outside the EU unless there is absolute necessity, requiring non-EU countries to provide "essentially equivalent" protection [2] - The EU Court has previously ruled that US laws do not provide "essentially equivalent" protection, yet the EU Commission approved the "Transatlantic Data Privacy Framework" (DPF) in July 2023, allowing EU companies to transfer data to US vendors despite existing surveillance laws [2][3] - The Privacy and Civil Liberties Oversight Board (PCLOB) is the only general oversight body in the US to ensure compliance with laws and commitments, but its effectiveness is questioned due to a lack of members [3] Group 2: Impact of the Trump Administration - The Trump administration has historically disregarded European concerns regarding data privacy, failing to support the PCLOB or the DPF's adequacy determination [4] - The administration's actions, including the dismissal of all three Democratic members of the PCLOB, have raised concerns about the board's ability to function effectively [5] - Trump's executive order aimed at ensuring accountability across federal agencies may undermine the Federal Trade Commission's ability to independently enforce DPF privacy principles [6] Group 3: Future Implications and Monitoring - The EU General Court's ruling does not preclude further challenges to the agreement, and MEP Latombe may appeal to the EU Court of Justice, which could have a different perspective [7] - If key elements relied upon by the EU are deemed ineffective, the EU may have to repeal the agreement, potentially leading to a situation similar to previous frameworks [7] - The US government's dual approach to cross-border data transfer is evident, as it takes extreme legal measures against TikTok while simultaneously protecting US companies from EU regulations [8]

Group 1 - The core point of the news is that FAW Jiefang Automotive Co., Ltd. has applied for a patent for a method and device related to vehicle operation condition design, which emphasizes data privacy protection during the design process [1] - The patent application was published under the number CN120524514A, with an application date of May 2025 [1] - The method involves obtaining vehicle operation parameters and merging data to determine a time series of operational parameters, which will be used as the design result for vehicle operation conditions [1] Group 2 - FAW Jiefang Automotive Co., Ltd. was established in 2002 and is located in Changchun City, primarily engaged in the automotive manufacturing industry [2] - The company has a registered capital of 1,080,301.25 million RMB and has invested in 23 enterprises [2] - FAW Jiefang has participated in 5,000 bidding projects and holds 5000 patent records along with 272 trademark records [2]

Core Viewpoint - The rapid development of big data and artificial intelligence has highlighted data security and privacy issues, with traditional data transmission methods posing significant risks. The introduction of blockchain technology offers new solutions, exemplified by MicroAlgorithm Technology's innovative BlockFL architecture, which ensures secure, efficient, and privacy-protecting data transmission [1][6]. Group 1: BlockFL Architecture - BlockFL architecture utilizes blockchain networks to achieve efficient data exchange and synchronization in federated learning, allowing devices to upload local model updates and download global model updates quickly and effectively [2]. - The decentralized nature and high concurrency of blockchain ensure that all devices receive the same global model updates, maintaining consistency and accuracy in model training [2]. Group 2: Process Overview - Initialization involves the system administrator creating an initial model and broadcasting it to all participating nodes while the blockchain records metadata of the federated learning activity [4]. - Each node trains the model on its local dataset without exposing original data, thus protecting data privacy [4]. - Nodes upload encrypted model parameters to the blockchain, where smart contracts validate their effectiveness and integrity, preventing malicious actions [4]. - Once verified, a central server or designated aggregation node extracts parameters from the blockchain, averages them, and generates a new version of the global model [4]. - The updated global model is then broadcasted to all nodes for the next training round, with the blockchain ensuring traceability of all operations [4]. - An incentive and penalty mechanism is integrated into BlockFL to encourage participation and quality data contribution, with smart contracts automatically executing rewards and penalties [4]. Group 3: Applications and Future Prospects - BlockFL architecture can be applied across various sectors, including healthcare, financial risk control, smart manufacturing, and smart cities, facilitating data collaboration while maintaining security and privacy [5]. - In healthcare, BlockFL enables hospitals to collaboratively train diagnostic models while protecting patient privacy; in finance, it allows institutions to identify fraud without sharing sensitive information; in smart manufacturing, it promotes collaboration between factories; and in smart cities, it supports inter-departmental cooperation without compromising sensitive data [5]. - The combination of blockchain and federated learning in BlockFL addresses traditional data transmission challenges, enhancing efficiency and accuracy in model training, positioning it as a significant technological support in data transmission and machine learning in the future [6].

Group 1 - OPPO showcased multiple cutting-edge AI terminal technologies and applications at the 2025 World Artificial Intelligence Conference (WAIC 2025) aimed at enhancing user work, life, and health management [1] - OPPO's Vice President and head of the Health Lab, Zeng Zijing, announced the industry's first AI health record and the OPPO AI Health Manager, which creates a comprehensive service system for health management, shifting from "passive treatment" to "active prevention" [3] - The AI health management system is designed to provide personalized and intelligent health management experiences for users, emphasizing the importance of understanding health concerns through AI interpretation of health reports [3] Group 2 - At WAIC, OPPO's General Manager of the Smart Assistant Department, Wan Yulong, revealed a collaborative development of edge AI parallel decoding acceleration technology, achieving over 8 times decoding acceleration on the latest chip platform, with peak output speed exceeding 200 tokens per second for 3B models [5] - This breakthrough in edge AI computing power lays a solid foundation for the continuous evolution of OPPO's system-level AI capabilities, transitioning mobile phones from "efficiency tools" to "efficient and dedicated smart assistants" [5] - OPPO emphasizes a "high-efficiency & dedicated" AI value proposition, focusing on a "1+3" core scenario strategy to enhance user experiences in various aspects of life, learning, and work [5] Group 3 - With the rise of AI, user demand for personal data privacy protection has significantly increased, making security and trust key factors in choosing AI products [7] - OPPO's Technical Strategy Planning Director, Chen Xiaochun, highlighted the importance of privacy protection, integrating it into the entire product design process, ensuring that personal information is processed only with user consent [7] - Looking ahead, OPPO aims to continue exploring innovative applications of AI technology while strictly protecting user privacy, focusing on multi-modal perception, intelligent agent development, and personalized experiences for users [7]