Group 1: Business Focus and Growth Strategies - The company is focusing on stable investment sectors such as telecommunications, finance, and energy for revenue growth [2] - In the cloud security sector, the company promotes the T-ONE CLOUD solution, addressing the needs of various clients including operators and government [3] - The company has served over 100 clients in cloud security, emphasizing a full-stack security product approach [3] Group 2: Market Opportunities and Product Development - The company is developing integrated security solutions for the rapidly growing Internet of Vehicles market, offering a comprehensive security service [3] - The AI+SOC product line, launched in 2024, has improved detection and response times by over 50% [4] - The ISOP security operation platform has seen steady revenue growth, with significant increases in SaaS-based security services [5] Group 3: Financial Goals and Performance Metrics - The company aims for a revenue target of at least 2.53 billion CNY and to achieve positive net profit by 2025 [5] - Specific strategies include focusing on stable customer demands in key industries and enhancing product competitiveness [5] Group 4: Competitive Advantages and Market Position - The company possesses leading self-research and technology advantages, catering to diverse enterprise-level client needs [6] - It primarily serves government, finance, telecommunications, energy, and education sectors, providing comprehensive security solutions [6] Group 5: Risk Management and Market Expansion - The company is actively exploring the network security insurance market, collaborating with major insurance firms for business expansion [9] - Plans for overseas market expansion focus on regions like ASEAN, Middle East, and Latin America, leveraging the "Belt and Road" initiative [10]

面对新兴挑战，行业正在探索新的防御范式。动态防御、AI对抗AI、隐私计算等新技术方向逐渐成为 共识。然而，技术只是解决方案的一部分。标准体系的缺失、跨厂商协作的障碍、法律伦理的模糊地 带，都是需要全社会共同应对的课题。 21世纪经济报道记者 董静怡 杭州报道 人工智能已经不再是一个独立的、孤立的技术概念，而是贯穿网络安全各个领域，AI智能体也已经从 实验室走入千家万户。 安恒信息CTO刘博向21世纪经济报道记者表示，在大模型时代，人们对安全的感知比以往更为强烈。一 方面，大模型涉及的数据量庞大且复杂；另一方面，在应用大模型时，往往需要处理大量敏感数据的交 互。无论是个人还是企业都对此更加重视。 然而，AI的发展催生了前所未有的安全挑战——攻击与防御的边界因技术迭代而模糊，传统规则式防 护体系在动态化、智能化的新型威胁前逐渐失效。 日前，在2025中国数谷·西湖论剑大会上，安恒信息董事长范渊指出，安全与智能密不可分，安全的边 界决定着智能的发展疆域，而智能的高度也将重塑安全的维度。 安全挑战加剧 大模型技术加速向金融、政务、企业等领域应用，多种安全风险相互交织、叠加放大。安恒信息首席安 全官、高级副总裁袁明坤向 ...

Investment Rating - The report does not explicitly provide an investment rating for the industry Core Insights - The emergence of AI technologies, particularly large language models (LLMs) and AI agents, is transforming the field of Offensive Security, including vulnerability assessments, penetration testing, and red teaming [23][24] - AI offers superior capabilities in Offensive Security, including data analysis, code and text generation, planning realistic attack scenarios, reasoning, and tool orchestration [24] - The integration of AI can enhance scalability, efficiency, and speed in identifying complex vulnerabilities, ultimately improving overall security posture [24][26] Summary by Sections Executive Summary - AI technologies are redefining Offensive Security as a multifunctional and powerful general technology [23] - The report explores the transformative potential of LLM-based AI in addressing current challenges in Offensive Security [23] Key Findings - Challenges in Offensive Security include a shortage of skilled professionals, increasingly complex environments, and the need to balance automated and manual testing [24] - AI can assist in automating reconnaissance, optimizing scanning processes, assessing vulnerabilities, generating comprehensive reports, and even autonomously exploiting vulnerabilities [24] - There is no single AI solution that can completely revolutionize Offensive Security; continuous experimentation is necessary to find effective solutions [24][25] Recommendations - AI should be integrated into automated tasks to enhance human capabilities in Offensive Security [26] - Human oversight is essential to verify AI outputs and improve quality [26] - Implementing a robust governance, risk, and compliance (GRC) framework is crucial for ensuring the safe and ethical use of AI [26] Offensive Security - Offensive Security involves proactively simulating attacker behavior to identify system vulnerabilities [27] - The effectiveness of Offensive Security is maximized when aligned with organizational goals and priorities [28] Current Challenges in Offensive Security - The expanding attack surface due to new technologies and remote work increases the difficulty of identifying potential entry points [34] - Advanced threats and the need for diverse assessment methods further complicate the landscape [34] - Balancing automation with manual testing remains a continuous challenge [34] AI in Offensive Security - AI encompasses a range of technologies aimed at simulating human intelligence, with a focus on LLMs [37] - LLMs can process and generate language, analyze large datasets, and assist in various tasks within Offensive Security [38] AI-Driven Offensive Security - AI technologies are opening new avenues in Offensive Security, enabling the simulation of advanced cyber attacks and the identification of vulnerabilities before malicious actors can exploit them [45] - AI can propose attack strategies, generate test cases, and learn from interactions to enhance its effectiveness [46] AI Enhancement and Autonomy - AI agents can autonomously or semi-autonomously execute Offensive Security tasks through iterative cycles of planning, reasoning, execution, and analysis [47] - The degree of autonomy granted to AI must be balanced with the risks associated with automation, especially in critical applications [47] Stages of Offensive Security - The report outlines five stages of Offensive Security: reconnaissance, scanning, vulnerability analysis, exploitation, and reporting [31] - AI can enhance human tasks across all stages, providing varying degrees of autonomy while ensuring compliance with regulations and organizational policies [50] Future of AI in Offensive Security - AI is expected to continue evolving towards higher levels of autonomy and automation, further enhancing the capabilities of Offensive Security [79] - The democratization of AI technology is lowering the barriers to entry for security testing, allowing more individuals and organizations to engage in robust security practices [80]

Core Insights - The conference "2025 China Data Valley · West Lake Sword Conference" emphasizes the importance of data flow for enabling various industries and the critical role of data in the AI revolution [2][6] - The lack of high-quality datasets is identified as a significant bottleneck in the deep integration of data and AI, which is essential for advancing from AI to AGI [2][3] Group 1: Data Flow and AI Integration - Data is considered the key element in the AI revolution, akin to fuel for a rocket, highlighting its necessity for intelligent systems [2] - The current limitations in data flow are attributed to security and quality issues, particularly concerning private data that is essential for model application in specific scenarios [3][5] - The industry is exploring synthetic data to address the scarcity of quality data, which affects model performance [3] Group 2: Trusted Data Space - The concept of a trusted data space is proposed as a solution to reconcile the development and utilization of data elements with security protection [6][10] - Trusted data spaces are designed to ensure secure and controllable data flow through a combination of hardware and software solutions, including privacy computing and blockchain technologies [10][11] - The national plan aims to establish over 100 trusted data spaces by 2028, with a focus on creating a comprehensive set of rules and standards for data security [11] Group 3: AI's Role in Data Security - The transition from static to dynamic data security is necessary to address the challenges in data flow, with AI playing a crucial role in this transformation [11][12] - AI technologies can enhance data classification and security by understanding the context and value of data, as well as identifying normal versus abnormal API usage [12] - The integration of AI in data security applications is seen as a significant opportunity for improving data management and protection [12]

分析师解释道："鉴于安全环境日益严峻，CrowdStrik已采取战略性措施，将人工智能整合到其产品组 合中，以满足对高级网络安全解决方案日益增长的需求，并保护整个技术堆栈中的IT和人工智能基础设 施。CrowdStrike仍然是网络安全领域的黄金标准，该公司正处于多年增长故事的早期阶段，预计其净 新增年度经常性收入将在下半年重新加速增长。我们相信，CrowdStrik在未来几年仍将处于确保人工智 能革命安全的前沿。" Wedbush Securities维持对网络安全公司CrowdStrike(CRWD.US)的"跑赢大盘"评级，并将目标价从395美 元上调至475美元。Wedbush表示，最近对CrowdStrike的实地检查结果"稳健"。 Wedbush分析师表示："我们将CrowdStrike的目标价从395美元上调至475美元，这反映了其网络平台业 务的强劲发展势头。我们最近的调查显示，这家网络安全巨头明年的势头非常强劲。CrowdStrike仍然 是我们最看好的科技公司之一，我们看到交易势头正在延续，人工智能也为这家占据优势地位的科技领 军企业带来了明显的利好。" 分析师表示，CrowdStrike ...

Group 1 - A Swiss delegation, consisting of over 25 representatives from academia and private institutions, arrived in Hong Kong on May 10 for a week-long visit aimed at promoting international partnerships and expanding bilateral exchanges, with a focus on understanding the innovation ecosystem in Greater China [1] - The first stop of the Swiss delegation was a seminar in Hong Kong on May 12, organized by the Swiss Digital Security Alliance, themed "Hong Kong - Gateway to China and Asia," which focused on digital transformation, cybersecurity, and cross-border investment in the era of artificial intelligence [1] - Hong Kong's Acting Director of Investment Promotion, Liu Zhi Yuan, highlighted the advantages of Hong Kong for international tech companies, emphasizing its potential as a partner for Switzerland in various sectors including life sciences, deep tech, fintech, education, and creative industries [2] Group 2 - Liu Zhi Yuan noted that the recent passage of the new Computer Systems Security Law enhances the protection of critical infrastructure in Hong Kong, allowing businesses to operate in a safer environment and boosting confidence [2] - The introduction of the Greater Bay Area standard contracts facilitates the secure and efficient transfer of personal data, promoting cross-border collaboration and seamless development of digital services [2] - Franz Grüter, Chairman of the Swiss Digital Security Alliance, stated that the visit aims to establish sustainable cooperative relationships, highlighting Hong Kong's strategic value as a connection point between international and Chinese markets for Swiss enterprises [2][3] Group 3 - The Hong Kong Digital Policy Office shared insights on the development of digital infrastructure and AI ecosystem policies, emphasizing the need for policy innovation alongside technological upgrades in the digital age [2] - The seminar included panel discussions led by industry experts from Thales, Swire Coca-Cola, and the Hong Kong-Shenzhen Innovation and Technology Park, focusing on cybersecurity challenges in the AI era [2] - The Hong Kong Investment Promotion Office plans to actively collaborate with the Swiss Consulate General, the Hong Kong Swiss Chamber of Commerce, and other relevant institutions to support Swiss enterprises in expanding their business in Hong Kong [3]

Meeting Overview - The annual shareholder meeting of Yuanjiang Shengbang Security Technology Group Co., Ltd. is scheduled for May 19, 2025 [1] - The meeting will follow the regulations set forth in the Company Law, Securities Law, and the company's articles of association [1][2] Attendance and Registration - Shareholders and their representatives must arrive at least 30 minutes before the meeting to complete registration [2] - Required documents include a securities account card, identification, and any necessary authorization letters [2][3] Meeting Procedures - The meeting will proceed according to the agenda outlined in the notice [2] - Shareholders have the right to speak, inquire, and vote during the meeting [2][3] - Each shareholder can speak or ask questions up to two times, with each speech limited to five minutes [3] Voting Process - Voting will be conducted through a combination of on-site and online methods [4] - Shareholders must express their opinions on the resolutions as "agree," "disagree," or "abstain" [4][8] Agenda Items - The meeting will cover several key resolutions, including: - Approval of the 2024 Board of Directors' work report [7] - Approval of the 2024 Supervisory Board's work report [9] - Approval of the 2024 financial settlement report, which indicates a net profit of RMB 1,628,506.74 [10] - Approval of the profit distribution plan, which states no cash dividends will be distributed due to negative cash flow [11][12] - Proposal to reappoint the accounting firm for the 2025 financial report [14][16] Financial Performance - The company reported a revenue of RMB 294 million for 2024, reflecting a 0.93% increase year-on-year [26] - The net profit attributable to the parent company was RMB 1.628 million [26] Strategic Focus - The company aims to enhance its competitive edge through increased investment in research and development, particularly in network security products [27] - The company is focusing on expanding its marketing network and upgrading its service systems to support business growth [12][27] Governance and Compliance - The company has established a robust governance structure, with regular meetings of the Board of Directors and Supervisory Board [30][34] - The independent directors have actively participated in meetings and provided oversight on company operations [30][36] Future Directions - The company plans to maintain a steady growth trajectory and ensure the effective execution of shareholder resolutions [32] - There is a commitment to enhancing transparency and investor relations management [32][33]

Core Viewpoint - The article emphasizes the importance of utilizing advanced technologies like AI and EDR systems for proactive cybersecurity measures, enabling companies to trace and analyze cyber threats effectively. Group 1: EDR Technology and Functionality - The RISING EDR system employs a "full-link tracking and tracing" feature that collects multi-dimensional data from terminal systems, allowing for intelligent analysis to identify abnormal behaviors and reconstruct the attack chain from initial intrusion to final objectives [1][2]. - Traditional security measures often focus on passive defense, but RISING EDR shifts this paradigm to active threat hunting, providing a comprehensive view of the attack process and pathways [2][3]. Group 2: Threat Analysis and Response - RISING EDR utilizes the ATT&CK framework to categorize network threats accurately, offering a detailed overview of attack dimensions and affected hosts, which aids in formulating security strategies [3]. - The system's ability to visualize the attack chain through various perspectives, such as relationship networks and timelines, enhances the understanding of the attack narrative [5][7]. Group 3: User Interaction and Customization - The integration of RGPT technology in RISING EDR allows users, even those without cybersecurity expertise, to interact naturally with the system for threat tracing and analysis [7][9]. - Companies can customize threat hunting rules within RISING EDR, enabling them to tailor security strategies to their specific needs and maintain control over their cybersecurity posture [10][12].

Group 1 - The 2025 China Digital Valley West Lake Forum opened on May 10, focusing on the integration of data element marketization and artificial intelligence, aiming to inject new momentum into the high-quality development of the digital economy [1][3] - Key discussions included the challenges of data elementization, the need for a data-centered organizational structure, and the establishment of clear legal frameworks for data assets [3][6] - The forum featured multiple sessions, including a main forum and parallel discussions on AI applications, data security, and the role of AI in digital transformation [3][4] Group 2 - A significant initiative was launched to promote the integration of data elements and artificial intelligence, emphasizing institutional, technological, organizational, and ecological innovations [6][9] - The forum highlighted the construction of a trusted data space for the marine economy, which has gathered 277 types of high-quality marine data resources and developed over 40 innovative application scenarios [7][9] - Discussions on the evolution of data security governance emphasized the need for a dynamic data security system to address the increasing scale and speed of data circulation [10][9] Group 3 - The importance of AI in enhancing data security and operational sustainability was underscored, with AI being positioned as a key solution to evolving security challenges [10][4] - The forum also addressed the governance challenges posed by rapid technological advancements, advocating for a balanced approach to innovation and security [10][4] - Strategic collaborations were initiated, including agreements between various institutions to strengthen the digital economy's development [6][9]

证券研究报告 公司研究 [Table_ReportType] 点评报告 [Table_StockAndRank] 深信服(688369.SH) 投资评级： 上次评级： [Table_A 庞倩倩 uthor计算机] 行业首席分析师 执业编号：S1500522110006 邮 箱： pangqianqian@cindasc.com 信达证券股份有限公司 CINDA SECURITIES CO.,LTD 北京市西城区宣武门西大街甲127号金隅大厦 B座 邮编：100031 [Table_Title] 业绩逐步恢复，AICP 平台升级打开云业务增长空间 ➢ 云业务订单带动收入增长，利润端同比亏损收窄。收入端：2025 年一 季度实现营收 12.62 亿元，同比增长 21.91%，主要是由于今年第一季 度公司云业务订单增速较好，带动公司整体收入实现增长，且公司去年 同期收入基数较低。利润端，公司 2025 年第一季度实现归母净利润- 2.50 亿元，亏损同比收窄 48.93%。此外，2025 年一季度公司经营性 净现金流同比增长 28.00%，主要是由一季度产品销售收到的现金增加 以及购买原材料支付的货款减少所致。 ➢ ...