就在公众愈发关注个人数据被互联网巨头白白拿去训练AI之际，一个名为Neon的应用却反其道而行之，通过协助消费者出售个人数据爆火。 截至发稿，Neon已经登顶美国App Store社交网络类别，同时在免费应用总榜上也冲到第三，仅次于谷歌Gemini和ChatGPT。 | | | 这款应用将自己标榜为一款通过AI产业赚钱的工具，声称通过录制并向"可信赖AI公司"出售录音，用户每年可以赚"数百或数千美元"。 当用户使用app打电话时，软件会自动录音并计费。如果双方都是Neon用户，软件会同时录制双方的语音，并按照30美分/分钟的价格支付费用。如果只有 用户是Neon用户，软件宣称只会录制用户的语音并按照15美分/分钟付费。单人单日最多可以通过打电话拿到30美元。 这样的一款应用能够在苹果商店上架，且能火起来，至少说明市场上的确有一部分人愿意用隐私来换取几美元现金——哪怕此举可能会带来不可知的沉重 代价。 如同市面上大多数互联网服务一样，Neon应用的风险也藏在服务条款的角落里。 Neon并未披露会把数据卖给哪些AI公司，而与诸多拥有用户数据的公司一样，Neon本身也存在数据泄露风险。 虽然公司宣称只会捕捉用户这一方 ...

一个传播视频信息的超级平台，遇上一个越来越抗拒全球化的世界。 本文首发于 2023 年 03 月 24 日 文 丨 高洪浩 实习生孙海宁 编辑 丨 黄俊杰 时隔三年，美国国会再次召开关于 TikTok 的听证会、TikTok 也再次面临分拆出售或被封禁的可能。 一位众议员在听证会上对接受质询的 TikTok CEO 周受资说，"我得感谢你，你做到了过去三四年难以 想象的事，你让共和党和民主党联合了起来。" 2019 年和 2020 年的两次听证会，TikTok 都选择不派人出席。但这一次 TikTok CEO 周受资到场接受 质询。 " 这 不是一个 好 兆 头 。 " 一位曾供职于美国智库与游说公司的人士在听证会召开前就对《晚点 LatePost》如此表示。按照过往经验，类似的听证会更多是议员们为自己发声创造的舞台。"TikTok 本 可以和前两次一样拒绝不出席，但这一次选择现身说明内部判断当前局势不容乐观，只能放手一搏。" 结果也是如此，近 5 个小时的听证会，周受资只有开场得到了近 6 分钟不被打断的时间完整陈述 TikTok 的立场。大部分时候，他说不到半分钟就会被打断，有时只能说 4-5 秒，就被要 ...

以下文章来源于时代财经APP ，作者冯恋阁 时代财经APP . 聚焦于企业财经新闻，互联网新闻信息服务许可证编号：44120230006 来源丨 时代财经APP（ID：tf-app） 作者丨 冯恋阁 图源丨Midjourney 2025年是当之无愧的Agent元年。 从Manus爆火，到各家终端厂商开始推出并迭代端侧AI助手，不难发现，厂商也意识到了这背后的机会。 与曾经的大语言模型不同，AI Agent能做的比聊天更多。 只需要一句自然语言指令，手机就能自动完成点外卖、发红包等曾经需要多步点击的任务。 大部分AI Agent能够代管手机，是使用了安卓系统里特殊的无障碍权限。这个本来是设计给视障人士的功能，能够让手机看到屏幕上的一切，同时，还可以 代替用户执行点击、滑动等操作。如今，厂商选择用这个权限来实现AI功能。 但当下，这份便利的背后实际隐藏着风险。在无障碍权限开放的背景下，AI应用基本"全知全能"，应用或许能够读取到包括支付密码、聊天记录在内的敏感 信息，也可能可以随意点击。 一个可见的事实是，随着AI Agent技术不断迭代，人类只会越来越习惯被硅基助手们包围的日常。 因此，当下更应该搞清楚AI A ...

Investment Rating - The report maintains a "Buy" rating for the company with a target price of HKD 20.9, up from a previous target of HKD 12.8 [1][4][7] Core Insights - The company reported a revenue of USD 938 million for the first half of 2025, representing a year-on-year growth of 47%, and a net profit of USD 32 million, which is a 3.4 times increase compared to the previous year [1][2] - The revenue from the advertising technology business reached USD 929 million, also growing by 47.6%, with the programmatic advertising platform Mintegral contributing USD 897 million, a growth of 48.6% [2][3] - The report highlights that the strengthening of antitrust policies is creating more opportunities for the industry, despite a slowdown in global economic growth [2] - The mixed monetization strategy is becoming mainstream, combining various revenue sources such as ads, subscriptions, and in-app purchases, which enhances revenue diversity [3] Financial Performance Summary - For 2025, the expected earnings per share (EPS) are projected at USD 0.06, with subsequent years forecasted at USD 0.08 for 2026 and USD 0.10 for 2027 [1][4] - The company’s revenue is expected to grow from USD 1,054.1 million in 2023 to USD 2,093.1 million in 2025, reflecting a growth rate of 38.8% [5][11] - Adjusted net profit is forecasted to increase significantly from USD 19.1 million in 2023 to USD 99.9 million in 2025, with a growth rate of 102.5% [5][11] Market Dynamics - The report notes that the advertising industry continues to grow despite challenges, driven by advancements in AI and machine learning technologies that enhance user targeting without compromising privacy [2][3] - The introduction of smart bidding products has significantly contributed to revenue growth, with over 80% of Mintegral's total revenue coming from this product line in the first half of 2025 [3]

Core Viewpoint - The launch of Apple's iPhone Air in China has been delayed due to regulatory issues surrounding eSIM technology, which is not yet supported by local telecom operators [2][3][4]. Group 1: iPhone Air Launch - The iPhone Air, Apple's lightest flagship smartphone, is currently not available for sale in China, with its release status marked as "to be updated" on official platforms [2]. - The iPhone Air only supports eSIM, which requires regulatory approval for telecom operators in China to offer eSIM services [2][3]. - Apple has indicated that the iPhone Air will only be supported by China Unicom for eSIM activation, requiring users to visit physical stores for identity verification [5][6]. Group 2: eSIM Technology and Regulatory Environment - eSIM technology is seen as a future trend in the telecom industry, but its adoption in China is hindered by regulatory and compliance challenges [3][11]. - Chinese telecom operators are prepared to offer eSIM services but are awaiting formal approval from regulatory bodies [4][11]. - The complexity of eSIM activation processes and the need for stringent security measures have led to delays in service rollout [8][9][14]. Group 3: Industry Perspectives - Industry analysts believe that the widespread adoption of eSIM in China is inevitable, but it will follow a cautious and phased approach [3][11]. - The telecom sector is currently focused on enhancing security measures for eSIM technology to address concerns over user privacy and fraud [8][9][14]. - Companies like China Unicom are exploring advanced technologies, such as blockchain, to improve the security of eSIM services [14].

【环球网科技综合报道】在AI技术迅猛发展的当下，隐私保护已日益成为大众关注的焦点，如何让用 户在享受技术带来便利的同时保障其隐私安全，成为各方尤其是各大科技企业面临的重要议题。 9月16日，在2025年国家网络安全宣传周期间，荣耀成为首批通过数安标准强基助力计划（DSEP） "个 人信息保护合规审计自审计能力评价"（简称"PCA-SC"）的企业，获得此次评价活动最高等级二星（规 范级）证书，是智能终端行业首家通过评价的企业。业内人士认为，此项证书不仅是对荣耀个人信息保 护合规审计能力的认可，更标志着荣耀在完善隐私保护体系、引领行业合规发展方面迈出了重要的一 步。 个人信息保护合规审计能力评价证书 荣耀终端股份有限公司： 感谢贵单位参与DSEP个人信息保护合规审计专题工作,依据 个人信息保护合规审计国家标准和实践指南，经评估，贵单位个 人信息保护合规审计自审计能力达到规范级,特发此证书。 针对AI换脸诈骗这一热点问题，荣耀设计开发"智能终端AIGC人脸合成反诈检测系统"，是全球首个基 于设备端AI的换脸反诈检测能力，能够在用户视频通话时对人脸进行检测与风险提示， 凭借出色的创 新性、先进性、实用性，此系统荣获2 ...

Core Viewpoint - The article highlights concerns regarding Microsoft's UCPD.sys, which allegedly contains hidden mechanisms that restrict Chinese software and potentially compromise user data security, raising alarms about privacy and national security [1][2][3]. Group 1: UCPD.sys Functionality - UCPD.sys is designed to protect user settings from being altered by third-party software, but it has been found to include a blacklist of numerous Chinese companies, effectively blocking their applications from modifying default settings [2][3]. - The program operates differently in various regions; in the EU, it complies with the Digital Markets Act (DMA) by allowing users to change default applications freely, while in China, it enforces restrictions that can revert settings to Microsoft’s defaults [3][4]. Group 2: Data Monitoring and Privacy Concerns - UCPD.sys actively monitors system changes and can execute hidden programs without user consent, raising significant privacy concerns as it may act as a backdoor for data collection [4][5]. - The driver collects detailed logs of user activity, including software usage and system modifications, which can be sent to Microsoft servers, potentially revealing sensitive information about user habits and preferences [5][6]. Group 3: Broader Implications and Historical Context - The article references past incidents where U.S. tech companies, including Microsoft, have been implicated in security breaches targeting Chinese entities, suggesting a pattern of behavior that could threaten critical infrastructure [6][7]. - The ongoing scrutiny of Microsoft’s practices has prompted calls for investigations and increased awareness among users regarding the risks associated with reliance on foreign software [7][8].

Core Viewpoint - The rapid development of intelligent connected vehicles in China is driven by strong government policies, but concerns about privacy and data security are emerging as significant issues [1][9]. Group 1: Policy and Market Trends - The State Council issued the "Opinions on Deepening the Implementation of 'Artificial Intelligence +'" in August, emphasizing the development of intelligent connected vehicles and the construction of a full-scenario intelligent interaction environment [1]. - The penetration rate of new intelligent connected passenger vehicles in China reached 68.2% in 2023, with expectations to rise to 78.9% by the end of 2025 [1]. Group 2: Privacy and Data Security Concerns - A data leak incident in 2023 affected over 2.15 million users, highlighting vulnerabilities in cloud storage systems [4]. - In 2024, privacy issues arose with high-profile brands like HiPhi and Li Auto, raising public concerns despite companies denying remote data access capabilities [1][4]. - Over 50% of popular models tested in 2024 were equipped with in-car cameras, but about 60% lacked adjustable physical covers, and 24% had cameras that could not be manually turned off, indicating inherent privacy risks [3]. Group 3: Legal and Regulatory Framework - China has established a legal framework centered around the Cybersecurity Law, Data Security Law, and Personal Information Protection Law, but the rapid evolution of technology often outpaces regulatory measures [6][7]. - A 2025 survey indicated that 71.8% of respondents believed that data from intelligent cabins should belong to users, reflecting a gap between consumer expectations and corporate privacy policies [7]. Group 4: Industry Self-Regulation and Consumer Awareness - The China Automobile Industry Association introduced a "Automobile Privacy Protection" label in November 2024, but only 56.6% of consumers actively inquired about data collection practices [7][8]. - Experts suggest that companies should adopt user-controllable designs and establish strict data management protocols to enhance privacy protection [8]. Group 5: Future Directions and Recommendations - Experts advocate for the adoption of advanced protective technologies, such as decentralized access control systems and privacy computing, to address privacy concerns [5][8]. - The industry is encouraged to create transparent authorization mechanisms and improve consumer awareness regarding data usage and privacy policies [8][9].

Group 1: Product Launch and Features - Apple introduced four new iPhone 17 models, with the iPhone Air being the thinnest at 5.6 mm and weighing 165 grams, starting at $999 [1][2] - The iPhone Air features a 6.5-inch screen and is designed to be lightweight, but it has limitations in battery life, requiring an additional battery accessory for extended use [1][10] - The iPhone Air lacks a 48MP ultra-wide camera, featuring only one rear camera, but introduces a new function allowing simultaneous capture from both front and rear cameras [10] Group 2: Market Position and Future Developments - Analysts speculate that the iPhone Air may serve as a precursor to Apple's upcoming foldable iPhone, expected to launch next year [2][11] - The foldable iPhone is anticipated to be Apple's high-end model, priced between $2000 and $2500, with a focus on minimal screen creasing [12] - Apple's competitors, such as Huawei and Samsung, have already released foldable devices, putting pressure on Apple to innovate in this segment [12][13] Group 3: AI and Software Developments - Apple is set to release iOS 26 in September, which will include new AI features supported by OpenAI's ChatGPT, although the rollout of Apple Smart features has been delayed [6][14] - The company has faced criticism for its slow progress in AI capabilities compared to competitors like Google and Meta, with many features still pending [17][20] - Apple's focus on privacy and security may limit the complexity of its AI functionalities, as the company prioritizes on-device processing over cloud services [21][22] Group 4: User Experience and Market Sentiment - Despite the lack of advanced AI features, iPhone users remain loyal, with satisfaction levels not significantly impacted by the absence of these capabilities [19][20] - Apple's strategy emphasizes enhancing user experience through privacy rather than leading in AI innovation, suggesting a different approach to technology integration [20][22]

Core Viewpoint - Apple is shifting its focus from cloud-based AI models to edge-based small models, exemplified by the release of FastVLM and MobileCLIP2, which prioritize speed and efficiency on personal devices [4][5][28]. Group 1: FastVLM Overview - FastVLM is a multimodal model capable of understanding both images and text, with a significant emphasis on speed, achieving response times 85 times faster than similar models [7][9]. - The model's architecture includes a new hybrid visual encoder, FastViTHD, which reduces the number of tokens generated from high-resolution images, thus improving processing speed without sacrificing accuracy [10][9]. - FastVLM is available in multiple sizes, including 0.5B, 1.5B, and 7B parameters, and can perform real-time tasks without cloud services [13][14]. Group 2: Apple's AI Strategy - Apple's AI strategy is divided into two plans: the "A Plan" focusing on large cloud models and the "B Plan" emphasizing small models for edge computing [32][36]. - The company has faced criticism for its slow progress in AI compared to competitors like Google and Microsoft, but it is now responding by investing heavily in AI initiatives and forming dedicated teams [33][36]. - Apple's commitment to privacy and user experience drives its focus on edge AI, ensuring that sensitive data remains on the device rather than being processed in the cloud [39][44]. Group 3: Market Context and Implications - The interest in small models is growing across the industry, but Apple's approach is unique in elevating it to a strategic priority for survival [50][51]. - The performance of small models can be optimized for specific tasks, making them suitable for applications in various sectors like healthcare and finance [48]. - Apple's hardware advancements, particularly in its A-series and M-series chips, provide a strong foundation for implementing efficient edge AI solutions [46][48].