Group 1 - The core viewpoint of the report indicates that AI models are increasingly infiltrating job-seeking safety across various channels, with 80% of accounts banned by the platform being proactively intercepted by AI-supported risk control systems [1][2] - The report highlights a new trend in illegal activities disguised as recruitment, characterized by "industrialization, gang formation, and outsourcing," prompting the platform to enhance AI's ability to identify violations and collaborate with law enforcement to combat crime [1] - The platform assisted in solving 58 cases and apprehending 607 suspects, successfully dismantling several criminal gangs involved in fraudulent recruitment schemes [1] Group 2 - To evade platform regulations, some violators intentionally lure job seekers away from the platform to third-party social media or offline interviews, revealing their true intentions [2] - In cases involving sexual misconduct, violators often shift to third-party platforms to make inappropriate requests or send explicit content, prompting the platform's safety team to optimize warning mechanisms [2] - The platform has significantly increased the dissemination of safety alerts and anti-fraud educational information to job seekers, pushing related information nearly 250 million times throughout the year [2]

Core Viewpoint - BOSS Zhipin-W (02076) plans to purchase a total of 315,200 Class A ordinary shares, equivalent to 157,600 American Depositary Shares, through internal resources on January 20, 2026, for the benefit of eligible participants [1] Group 1 - The company will execute the share purchase in the open market [1] - The shares will be held in trust for qualified participants [1]

Core Viewpoint - BOSS Zhipin (02076) announced a plan to purchase a total of 315,200 Class A ordinary shares, equivalent to 157,600 American Depositary Shares, to be held in trust for eligible participants [1] Group 1 - The purchase is scheduled to take place on January 20, 2026 [1] - The shares will be acquired in the open market based on the terms and conditions of the plan [1] - The funding for the purchase will come from internal resources of the company [1]

Core Viewpoint - BOSS Zhipin-W (02076.HK) has announced a share repurchase plan that was conditionally approved on December 14, 2022, which will take effect following the listing of its Class A ordinary shares on the Hong Kong Stock Exchange on December 22, 2022 [1] Group 1 - The company plans to repurchase a total of 315,230 Class A ordinary shares, equivalent to 157,615 American Depositary Shares (ADS), using internal resources [1] - The repurchase will be executed by a trustee in the open market, in accordance with the terms and conditions of the plan [1] - The shares will be held in trust for eligible participants as defined in the plan [1]

香港交易及結算所有限公司及香港聯合交易所有限公司對本公告之內容概不負責，對其準確性 或完整性亦不發表任何聲明，並明確表示概不就因本公告全部或任何部份內容而產生或因倚賴 該等內容而引致之任何損失承擔任何責任。 香港，2026年1月21日 謹此提述看準科技有限公司（「本公司」）於2022年12月14日有條件批准及採納的 首次公開發售後股份計劃（「計劃」），該計劃於2022年12月22日本公司A類普通 股（「A類普通股」）在香港聯合交易所有限公司主板上市後生效。於2026年1月20 日，由本公司內部資源撥付，計劃受託人（「受託人」）根據計劃之條款及條件在公 開市場上購買合共315,230股A類普通股（以157,615股美國存託股的形式），並以 信託方式為合資格參與者（定義見計劃）持有該等股份。受託人所購買並以信託方 式持有的A類普通股詳情如下： 購買日期： 2026年1月20日 所購買A類普通股總數： 315,230股 於2026年1月20日，所購買A類普通股佔本公司已發行及 流通在外股份總數的百分比： 約0.03% 所購買每股A類普通股的平均代價（不包括所有相關開支）： 9.49美元 緊隨購買後受託人所持A類普通 ...

Core Insights - The report by BOSS Zhipin highlights the increasing penetration of AI large models in job-seeking safety, with 80% of accounts banned being proactively intercepted by AI-supported risk control systems [2] - The report identifies a new trend in illegal activities under the guise of recruitment, characterized by industrialization, gang formation, and outsourcing [2] - BOSS Zhipin's security team has collaborated with law enforcement to crack down on criminal activities, assisting in the resolution of 58 cases and the arrest of 607 suspects [2] Group 1: AI's Role in Combating Black and Gray Industries - The report indicates that the characteristics of black and gray industries in recruitment are becoming increasingly professionalized, with tactics such as crowdsourcing to induce HR to sell resumes and using "resume tool people" [3] - The security team utilizes AI large models for pattern recognition to identify and ban accounts involved in these activities, while also providing leads to law enforcement [3] - Feedback from law enforcement indicates that several individuals are under investigation for their involvement in these activities [3] Group 2: AI Applications in Risk Control - The report reveals two core applications of AI in job recruitment risk control: semantic recognition and behavior modeling [5] - AI's deep semantic inspection capabilities allow the identification of subtle violations in job descriptions, significantly improving interception efficiency [5] - The security team has launched multiple governance initiatives targeting issues such as sexual harassment in recruitment and fraudulent job postings, successfully banning tens of thousands of accounts [5] Group 3: Safety Measures for Job Seekers - BOSS Zhipin has enhanced its safety alerts and anti-fraud educational information, sending nearly 250 million safety tips to job seekers throughout the year [6] - The security team has optimized alerts for job seekers when suspicious behavior is detected, such as recruiters frequently encouraging users to move to third-party platforms [6] - Job seekers are advised to maintain communication within the platform and report any suspicious activities immediately [6] Group 4: Comprehensive Safety Strategy - BOSS Zhipin has segmented its job-seeking safety protection into three areas: core protection zone, risk buffer zone, and ecological co-governance zone [7] - The core protection zone involves an "artificial + AI" review system covering the entire recruitment process to ensure authenticity [8] - In collaboration with law enforcement and industry partners, BOSS Zhipin aims to reduce the survival space for illegal recruiters and promote transparency in job postings [8]

Core Insights - The report by BOSS Zhipin highlights the increasing integration of AI in job safety governance, with 80% of accounts banned being proactively intercepted by AI-supported risk control systems [1] - The report identifies a new trend in illegal activities under the guise of recruitment, characterized by industrialization, gang formation, and outsourcing [1] - BOSS Zhipin's safety team has collaborated with law enforcement to crack down on criminal activities, resulting in the resolution of 58 cases and the arrest of 607 suspects [1] AI's Role in Combating Black and Gray Industries - The report indicates that black and gray industries in recruitment are becoming more specialized, with tactics such as crowdsourcing to induce HR to sell resumes and using "resume tool people" [2] - The safety team employs AI models to identify and ban accounts involved in suspicious activities, while also providing leads to law enforcement, resulting in several individuals being investigated [2] AI Applications in Risk Control - The report outlines two core applications of AI in job recruitment risk control: semantic recognition and behavior modeling, enhancing the ability to detect subtle violations in job descriptions [4] - AI has improved the efficiency of identifying various types of violations, allowing for quicker detection and banning of problematic accounts [4] Governance Initiatives - In 2025, with AI support, BOSS Zhipin launched multiple initiatives to combat issues such as sexual harassment in recruitment, illegal fees, and fraudulent activities, leading to the banning of tens of thousands of accounts [5] Safety Alerts and User Education - The report reveals that BOSS Zhipin sent nearly 250 million safety alerts to job seekers, aiming to build a "first line of defense" against recruitment fraud [6] - The safety team has optimized alerts for job seekers, advising them to maintain communication within the platform and report any suspicious activities [6] Three-Tiered Safety Framework - BOSS Zhipin has established a three-tiered safety framework: core protection zone, risk buffer zone, and ecological co-governance zone, collaborating with various stakeholders to enhance job safety [7] - The core protection zone involves an "artificial + AI" review system covering the entire recruitment process to ensure authenticity [8] - In the ecological co-governance zone, BOSS Zhipin has partnered with law enforcement and industry players to reduce the presence of illegal recruiters, including signing self-regulatory agreements with delivery companies [8]

报告指出，以招聘为名开展违规违法活动的黑灰产呈现"工业化、团伙化、外包化"的新趋势。对应平台 在加强AI对违规行为识别力度的基础上，配合警方从源头打击犯罪，协助警方破获案件58起，抓获嫌 疑人607人，成功打掉数个"套路运""招转培"犯罪团伙。 同时，安全团队也加强了向求职者发送安全提示与防骗科普信息的力度，在求职者心中构建"第一道防 线"，全年向求职者推送相关信息近2.5亿次。 同时，安全团队也加强了向求职者发送安全提示与防骗科普信息的力度，在求职者心中构建"第一道防 线"，全年向求职者推送相关信息近2.5亿次。 责任编辑：何俊熹 新浪科技讯 1月20日晚间消息，BOSS直聘发布《2025年安全治理报告——在动态博弈中凝聚盟友》。 报告显示，AI大模型正在渗透求职安全的各个链路，平台全年封禁处置的账号中，80%由AI支持的风控 系统主动拦截。 新浪科技讯 1月20日晚间消息，BOSS直聘发布《2025年安全治理报告——在动态博弈中凝聚盟友》。 报告显示，AI大模型正在渗透求职安全的各个链路，平台全年封禁处置的账号中，80%由AI支持的风控 系统主动拦截。 报告指出，以招聘为名开展违规违法活动的黑灰产呈现"工 ...

Core Viewpoint - The article discusses the competitive landscape of AI companies, focusing on two emerging players, MiniMax and Zhipu AI, highlighting their different strategies and market positions in the rapidly evolving AI sector [6][21]. Group 1: MiniMax - MiniMax is characterized as an aggressive player focusing on C-end driven, multi-model approaches, starting with a product aimed at AI virtual social interactions [7][8]. - The company’s first product, Glow, launched in October 2022, laid the foundation for its focus on emotional interaction, leading to significant revenue contributions from its later products, Talkie and Xingye, which accounted for 63.7% of MiniMax's revenue in 2024 [8][10]. - By 2025, MiniMax expanded its product offerings to include independent models for text, video, and voice, creating a diversified product matrix supported by multi-modal technology [9][10]. - MiniMax's revenue structure is promising, with over 70% of its income coming from overseas, primarily from C-end user subscriptions, indicating strong growth potential [11]. - Despite its growth, MiniMax faces challenges, including a low market share of 0.3% in the global AI market as of Q3 2025, and the intense competition from larger players requiring significant investment in computing power and data [11][12]. Group 2: Zhipu AI - Zhipu AI, founded in 2019 and rooted in academic research, has a more traditional approach, focusing on B-end services and a unified large model strategy [13][14]. - The company has undergone 18 rounds of financing before its IPO, attracting significant investment from various industry giants, which has positioned it favorably in the market [15]. - Over 80% of Zhipu AI's revenue comes from local enterprises, with a high client concentration, where the top five clients contributed approximately 40% of its revenue in the first half of 2025 [16]. - Zhipu AI maintains a high gross margin of over 50%, but its growth potential may be limited compared to MiniMax, which explains the disparity in stock performance post-IPO [16][19]. Group 3: Market Outlook - The article emphasizes the divergent paths of MiniMax and Zhipu AI, showcasing the varied strategies within the AI industry, from project-based to product-based models, and the focus on domestic versus international markets [21][22]. - The global AI landscape is evolving, with significant investments and developments anticipated, including OpenAI's potential IPO and the ongoing competition among major players [23]. - A report from McKinsey indicates that 88% of organizations are using at least one AI tool, suggesting a growing trend towards AI integration across industries, which will ultimately reward those who effectively embed AI into their operations [23].

Core Insights - Huaying Capital, founded in 2008, is a leading private equity investment institution in China, focusing on innovation-driven investments with a core emphasis on digitalization [1] - The company has developed a dual-currency management system, managing over 13 billion RMB, primarily investing in A and B rounds while also engaging in early-stage and later-stage investments [1] - Huaying Capital has invested in over 300 projects across key sectors such as technology, consumer, enterprise services, and digital economy, supporting both mature listed companies and high-potential innovative firms [1] Investment Strategy - The company aligns with the national strategy of "new productive forces," focusing on systematic and forward-looking investments in core areas like artificial intelligence, semiconductor chips, and aerospace economy [2] - Huaying Capital has established an investment matrix covering critical tracks of new productive forces, with a comprehensive investment approach in the AI and embodied intelligence sectors [2] - In the semiconductor field, the company has invested in key segments from design to equipment, supporting domestic chip development and breaking technological barriers [2] Future Outlook - The long-term investment strategy centered on digitalization and new productive forces has created a competitive advantage for Huaying Capital, providing a replicable path for the venture capital industry to support national strategies and empower the real economy [3] - The company aims to continue deepening its focus on innovative sectors, providing long-term capital to foster the growth of technology innovation enterprises and contribute to the development of globally competitive Chinese firms [3]