韩国科技信息通信部周二公布的数据显示，2025年韩国企业在数据保护上的支出较上年增长了14.3%， 今年企业面临的客户信息盗窃威胁有所增加。 数据显示，韩国企业今年在相关领域的支出为2.4万亿韩元（约合16.7亿美元），而2024年的支出为2.1 万亿韩元。调查结果基于对773家公司提交的监管文件的分析。 据韩国科技信息通信部称，今年专门从事数据保护的员工人数为8506人，比2024年的7681人增加了 10.7%。 该部在一份新闻稿中表示："数据保护相关支出和相关员工数量的增加同时表明，企业对该领域的兴趣 日益浓厚。" 按行业分类，金融保险公司的数据保护平均支出为85亿韩元，信息通信公司为62亿韩元。 韩国科技信息通信部周二公布的数据显示，2025年韩国企业在数据保护上的支出较上年增长了14.3%， 今年企业面临的客户信息盗窃威胁有所增加。 数据显示，韩国企业今年在相关领域的支出为2.4万亿韩元（约合16.7亿美元），而2024年的支出为2.1 万亿韩元。调查结果基于对773家公司提交的监管文件的分析。 据韩国科技信息通信部称，今年专门从事数据保护的员工人数为8506人，比2024年的7681人增加了 10 ...

AI Dynamics - The prospectus of Zhipu, the first domestic large model company, reveals significant financial losses, with net losses of 143 million yuan in 2022, 788 million yuan in 2023, and projected losses of 2.958 billion yuan in 2024, totaling over 6.2 billion yuan by mid-2025 [1][2] - The prospectus highlights the importance of low hallucination rates as a key indicator of model reliability, with Zhipu's GLM-45 ranking second globally and lowest in China for hallucination rates by September 2025 [1][2] - Regulatory changes and compliance issues are emphasized as potential risks, with the prospectus noting that future laws may impose additional obligations that could adversely affect business operations and financial performance [2][3] - User data compliance is a critical concern, as the prospectus acknowledges the uncertainty surrounding the interpretation and application of data protection laws, which may impact the company's ability to safeguard user data [3][4] - Training data compliance is identified as an unstable factor, with the prospectus indicating that Zhipu may source training data from third-party vendors and public datasets, raising concerns about the legality and compliance of such data [4][5] - The potential misuse of AI technology by users poses a reputational risk to the company, as any negative outcomes could significantly impact business performance and financial outlook [5] Platform Governance - The newly issued "Internet Platform Pricing Behavior Rules" prohibit platforms from forcing or indirectly compelling operators to adopt automatic pricing systems or engage in unreasonable pricing practices [7][8] - TikTok has announced the establishment of a new joint venture in the U.S. focused on data protection, algorithm security, content review, and software assurance, named "TikTok USDS Joint Venture LLC" [8] - The Supreme People's Court has updated the civil case categories to include data and virtual property disputes, reflecting a growing focus on legal frameworks surrounding digital assets and intellectual property [9]

以下文章来源于算法爱好者 ，作者小蒜 TikTok 母公司字节跳动与三家牵头投资者签署协议，宣布成立美国合资新实体 TikTok USDS Joint Venture LLC，以出售逾 80% 美国资产的方式规避美方禁令，这场持续数年的业务风波终现落定迹 象。 根据备忘录披露的核心细节，新实体股权结构已明确划分： 值得注意的是，甲骨文、银湖资本与 MGX 三家将共同持有新实体 45% 股份，这一比例与路透社等媒 体 9 月的报道完全吻合。 TikTok 新实体将负责 TT 美国的数据保护、算法安全、内容审核和软件保障。 由字节全资控股的 TT 在美国其他实体，将继续负责电商、广告、市场运营等商业活动。 算法爱好者 . 算法是程序员的内功！「算法爱好者」专注分享算法相关文章、工具资源和算法题，帮程序员修炼内 功。 12 月 18 日，路透社和彭博社等外媒报道，TikTok CEO 周受资发出内部信，更新了 TikTok 美国业务 进展。 TikTok 在备忘录中表示， 一个由七人组成的董事会将负责监管这家新实体，其中多数成员是美国人。 交易完成后，美国 1.7 亿用户可继续正常使用平台服务。 据悉，新实体相关交 ...

（原标题：首届巴西—印度数字对话在巴西利亚举行） 巴西媒体11月26日综合报道，首届巴西—印度数字对话在巴西利亚召开，两国政府各部门官员、专 家和代表出席，体现双方在数字领域合作全面性和战略性。双方讨论了数字化转型的关键议题，包括人 工智能治理、公共数字基础设施、有效互联互通、高性能计算、半导体、科技创新、数据保护以及多边 场合协作等。双方将扩大双边协调，通过联合倡议、技术交流和定期磋商等方式在数字问题上建立更紧 密伙伴关系。双方重申，数字化转型必须建立在技术主权和包容性可持续发展的基础上。巴西欢迎印度 于 2026年2月在新德里主办人工智能影响力峰会（India AI Impact Summit 2026），届时巴方将派代表参 加。 ...

Core Points - Reddit has filed a lawsuit against AI startup Perplexity and three other companies for illegally scraping its data to train an AI-based search engine [1][2] - The lawsuit highlights the ongoing legal battles between content owners and tech companies over the use of copyrighted material for AI training [1] - Reddit claims that Perplexity's data usage has increased fortyfold since a cease-and-desist letter was sent last year [2] Group 1 - Reddit accuses Perplexity and other companies of bypassing data protection measures to steal data necessary for its "answer engine" system [1] - The lawsuit is part of a broader trend where content owners are taking legal action against companies that misuse their data for AI training [1] - Reddit's Chief Legal Officer stated that AI companies are engaged in a "data laundering" economy to acquire quality human content [1] Group 2 - Reddit claims that companies like Oxylabs, AWMProxy, and SerpApi have scraped its data without permission [2] - SerpApi and Oxylabs have expressed their intention to vigorously defend against Reddit's allegations [2] - Reddit is seeking unspecified monetary damages and an injunction to prevent Perplexity from using its data [2]

Group 1 - A data company utilized its cloud computing system to analyze personal information and internet behavior of users, generating user tags for targeted advertising, resulting in over 4.5 million yuan in revenue from January to May 2018 [1] - The company faced a fine of 1.5 million yuan from the Hangzhou Market Supervision Administration for violating the Anti-Unfair Competition Law by engaging in unauthorized advertising practices [1][2] - The revised Anti-Unfair Competition Law, which emphasizes data protection, officially took effect recently, with Hangzhou's regulatory system recognized for its leading role in data protection enforcement [5] Group 2 - A software technology company developed two "data migration" apps that allowed users to scrape detailed product data from third-party e-commerce platforms without consent, generating over 2 million yuan in service fees [3][4] - The company was fined 1.2 million yuan for disrupting market order and violating the Anti-Unfair Competition Law by using unauthorized data scraping techniques [4] - The Hangzhou Market Supervision Administration has identified and published a list of significant data protection enforcement cases, including those related to data scraping and commercial secrets [6]

Core Points - The 2025 China-Africa Internet Development and Cooperation Forum opened in Xiamen, focusing on building a community of shared future in cyberspace between China and Africa [1] - China proposed the "Action Plan for Building a Community of Shared Future in Cyberspace between China and Africa (2025-2026)" and established the "China-Africa Internet Enterprise Cooperation Network" [1] - The forum featured four sub-forums addressing digital divide, digital transformation, artificial intelligence development and governance, cybersecurity and data governance, and media cooperation [1] Industry Insights - Approximately 400 representatives from Chinese and 32 African countries attended the forum, including government agencies, universities, think tanks, and media [1] - African representatives expressed high appreciation and positive responses to China's cooperation initiatives, indicating alignment with their digital transformation needs [1] - The forum was co-hosted by the National Internet Information Office and the Fujian Provincial Government, with support from the Xiamen Municipal Government [1]

Core Viewpoint - The article highlights Ant Group's participation in the 2025 National Cybersecurity Publicity Week, showcasing its innovations in AI security, data protection, and intelligent risk control, particularly through the introduction of the gPass framework for AI glasses [1][2]. Group 1: gPass Framework - gPass is designed to provide a secure, interactive, and connected experience for AI glasses, addressing challenges such as fragmented ecosystems and limited application scenarios in the AI glasses industry [1][2]. - The framework employs technologies like trusted identity circulation, end-to-end encryption, and device authentication to ensure user information security and privacy [2]. - gPass has already partnered with brands like Rokid, Xiaomi, Quark, and Thunderbird, with plans to expand its applications to various life scenarios, including healthcare and travel [2]. Group 2: Advanced Security Technologies - Ant Group has introduced several advanced security technologies, including the ASL initiative for agent collaboration security and the "Ant Tianjian" model security solution, which includes features for detecting misuse and ensuring data privacy [3]. - The ZOLOZ Deeper technology effectively addresses threats from deep forgery, such as fake faces and voice synthesis [3]. - The "Trusted Data Space" product under Ant Group's Mican provides high-performance, low-cost, and secure data fusion capabilities, supporting various sectors [3]. Group 3: Risk Control Capabilities - Ant Group's financial technology division has demonstrated advanced risk control capabilities against document and voice forgery, achieving a 98% accuracy rate in fake document detection and covering over 50 types of voice synthesis [4]. - The company has collaborated with judicial authorities to address illegal financial intermediaries, involving over 200 individuals since 2024 [4]. - Ant Group aims to build a trustworthy AI governance system to ensure the authenticity and reliability of AI-generated content and agent behavior [4]. Group 4: Recognition and Awards - Ant Group's security technology has received multiple awards for its research and application in business security, AI security, and content security, including first prizes from various technology advancement awards [5].

Core Insights - The project focuses on the development of an AI-based disaster recovery backup system that offers users a simple, cost-effective, and integrated recovery solution [1][3] - The system provides automated and intelligent data backup, recovery, monitoring, and management, ensuring business continuity in the event of data loss or disasters [1][2] Group 1: Technology and Innovation - The AI-driven disaster recovery system addresses traditional bandwidth costs and manual switching issues, utilizing AI technologies for unified management and on-demand bandwidth allocation, significantly reducing costs [3] - The project introduces a revolutionary breakthrough in data protection, innovating the business model by offering SaaS services that charge annual service fees, particularly benefiting small and medium-sized enterprises [3][4] Group 2: Cost Efficiency - The application of this technology effectively reduces data storage scale and controls the exponential growth of full backup data in archiving systems, leading to significant savings in data storage space and data center operational costs [4] - AI contributes to resource efficiency and cost savings, including reductions in data center power consumption, cooling costs, and physical space requirements, as well as decreased storage capacity, network bandwidth, and IT personnel needs [4][5] Group 3: Service Improvement - AI significantly enhances backup performance, allowing for completion within limited backup time windows, and improves recovery performance by leveraging random access disk storage compared to sequential access methods [5][6] - The economic viability of disk-based backups is improved for a wider range of applications due to the advancements brought by AI [6]

Core Points - The French National Commission on Informatics and Liberty (CNIL) has fined Google a total of €325 million (approximately ¥2.7 billion) for displaying ads and inserting cookies in Gmail without obtaining valid user consent [1][3] - Google has six months to rectify the situation, or it will face an additional daily fine of €100,000 [1][3] Summary by Sections CNIL's Findings - CNIL's investigation revealed that Google Ireland Limited and Google LLC inserted ads into the "Promotions" and "Social" tabs of Gmail without user consent, violating the French Postal and Electronic Communications Code [1][3] - The commission found that users were misled into selecting cookies related to personalized ads during account creation, without clear information that consent was required for cookie storage and advertising [3] Penalties and Requirements - The total fine consists of €200 million for Google LLC and €125 million for Google Ireland Limited [3] - CNIL mandates that Google must stop inserting ads in Gmail without user consent and ensure valid consent for advertising cookies during account creation within six months [3] User Impact - The illegal cookie practices affected over 74 million users in France, with 53 million having seen ads in their Gmail [3] Google's Response - Google stated it is evaluating the CNIL's decision and emphasized that users have control over ad displays in its products [4] - The company has previously made adjustments in response to regulatory demands, including allowing users to opt-out of personalized ads [4] Historical Context - This is not the first penalty for Google; it has faced fines from CNIL for similar violations in 2020 and 2021, amounting to €100 million and €150 million respectively [4]