21世纪经济报道记者 吴佳楠 深圳报道 9月15日，粤港澳大湾区生成式人工智能安全发展联合实验室揭牌仪式在河套深港科技创新合作区举行。 联合实验室汇聚多方力量共同建设，参与单位包括粤港澳有关部门，腾讯、OPPO、小鹏汽车、深信服、视源、云蝶、云天励 飞、元象、深译、陆兮等AI应用龙头企业和AI专精特新企业，中山大学、广州大学、工信部电子五所等高校和科研院所。 联合实验室有关负责人向记者表示，联合实验室立足粤港澳大湾区科技创新高地，通过整合政府、企业、高校、科研机构及媒 体资源，构建"政产学研用"深度融合的创新生态，本质上是一种为适应人工智能时代而构建的"动态敏捷、多元协同"的创新治 理联合体。 聚焦三大领域发挥优势 当前随着人工智能技术向千行百业深度渗透，人工智能安全风险呈现跨领域、复合型、动态化特征，传统"事后监管"模式已难 以应对复杂挑战。 从技术层面看，生成式人工智能的随机性输出导致传统安全防护失效，传统规则引擎无法覆盖大模型的涌现风险，比如深度伪 造内容、隐蔽性数据泄露等；从生态层面看，生成式人工智能的对抗样本攻击、越狱攻击等新型威胁手段迭代加速，而多数市 场主体尤其是中小微企业的安全防御技术储备不足 ...

生成式人工智能是引领新一代科技革命和产业革命的核心驱动力，是加快培育和发展新质生产力的重要 引擎，为经济高质量发展注入新动能，与此同时，各类难以预知的风险和挑战也伴生而来。 要通过联合实验室整合数据 南方都市报（以下简称南都）：您是世界数据科学与分析领域的领军学者，能否结合数据领域深入研 究，谈谈对粤港澳大湾区生成式人工智能安全发展联合实验室发挥自身优势，更好服务人工智能安全发 展的期望？ 陈雷：粤港澳大湾区制造业很强，要把人工智能运用到传统行业，数据非常重要。通过实验室联合大湾 区各类高校，汇聚所有数据，做成大数据平台，供大家使用，做相应大模型测试。或者由联合实验室推 出数据测试平台，让各类大模型通过平台测试性能，找出不足，加以改进。 南都：如何确保数据质量，联合实验室能做些什么？ 陈雷：数据质量是全世界都想解决的问题，首要的问题是获取到的数据是不是有用？而且数据量要大， 量不大就没有所谓数据质量。 数据质量包括无监督、有监督。无监督的数据质量就是通过无监督的机器学习，发现一些质量偏差较远 的数据，或者叫异常数据分析。有监督的数据质量一般需要找专家或专业人士对数据打标签，然后用机 器学习的方法找出那些较差的 ...

Investment Rating - The report indicates a positive outlook for the life sciences industry, emphasizing the competitive advantage gained through artificial intelligence (AI) adoption [10][11]. Core Insights - The life sciences sector is leading in AI application, with 86% of companies believing they can embrace AI for competitive advantage, and 97% have already improved operations through AI [10][11]. - Despite the potential, many companies face challenges in achieving high returns on AI investments, with a significant portion only reaching break-even or low returns [7][8]. - The report outlines a structured framework for AI transformation in three phases: empowering employees, integrating AI into workflows, and evolving operational models [21][59]. Summary by Sections Introduction - The introduction highlights the transformative potential of AI in the life sciences industry, emphasizing the need for innovation in operations and value creation [16][17]. Overview - AI is recognized as a significant competitive advantage, with initial implementation results being encouraging [10][11]. - A well-adapted organizational structure is linked to higher investment returns [10]. Research Findings - The report reveals that 73% of companies have improved efficiency through AI, while 39% have enhanced financial performance [67]. - Data issues, including silos and quality concerns, are identified as major challenges in AI implementation [32][40]. Building Intelligent Life Sciences Enterprises - The report discusses the importance of integrating AI into daily operations and the need for a mixed organizational structure to drive innovation [24][41]. - Companies are encouraged to develop a culture that supports continuous learning and collaboration to maximize AI's potential [48]. Phase One: Empowering Employees - In this phase, companies focus on identifying areas where AI can automate tasks and improve workflows [66]. - Nearly three-quarters of respondents reported efficiency gains from AI, with a significant number also noting improvements in financial health [67]. Phase Two: Integrating AI into Workflows - Companies are advised to embed AI into various functions, enhancing operational efficiency and decision-making processes [61][64]. - The integration of AI should be aligned with business objectives to ensure strategic relevance [73]. Phase Three: Evolving Operational Models - The final phase emphasizes the need for companies to adapt their business models and ecosystems to leverage AI effectively [61][62]. - Organizations should focus on building trust in AI systems and ensuring compliance with ethical standards [48]. Key Recommendations - The report suggests that life sciences companies should prioritize developing a comprehensive AI strategy that aligns with business goals and stakeholder needs [48]. - Establishing a flexible and scalable technology infrastructure is crucial for maximizing AI's long-term value [48]. Conclusion - The life sciences industry is positioned to harness AI for significant advancements, but companies must address data challenges and cultivate a supportive culture to fully realize AI's benefits [42][43].

没有网络安全就没有国家安全，就没有经济社会稳定运行，广大人民群众的利益也难以得到保障。近年 来，我省深入贯彻落实习近平总书记关于网络强国的重要思想，网络安全政策法规逐步健全，工作机制 不断完善，防护能力显著增强，教育技术产业融合发展，广大人民群众在网络空间的获得感、幸福感、 安全感不断提升。 9月15日，2025年国家网络安全宣传周河南省活动主题论坛在商丘市举行，各方来宾齐聚一堂，关注技 术发展新趋势，探索如何做好网络安全工作，推动全社会共同营造清朗、健康的网络生态。 网络无远弗届，安全如影随形。当前，新一轮科技革命和产业变革加速演进，网络空间与现实世界深度 融合。统计数据显示，截至今年6月，我国网民规模达11.23亿人，互联网普及率达79.7%，数字红利惠 及亿万民众，网络空间已成为亿万民众共同的精神家园。互联网健康有序发展，不仅关系到国家安全和 社会稳定，更关乎亿万网民的切身利益。人民群众对提升互联网治理能力的要求越来越迫切，加强网络 空间法治建设成为必然选择。 网络安全为人民，网络安全靠人民。强化风险意识，树立正确网络安全观，守牢安全底线、筑牢安全屏 障；坚持依法治网，突出防护重点，加大网络和数据安全执 ...

工业和信息化部电子第五研究所副所长、联合实验室专家王蕴辉 开篇语 生成式人工智能是引领新一代科技革命和产业革命的核心驱动力，是加快培育和发展新质生产力的重要 引擎，为经济高质量发展注入新动能，与此同时，各类难以预知的风险和挑战也伴生而来。 安全是发展的基石，为进一步创新筑牢根基。2025年9月15日，粤港澳大湾区生成式人工智能安全发展 联合实验室揭牌成立。其将构建"政产学研用"深度融合的创新生态，致力服务企业发展、推动产业落 地、加强安全监管，努力实现属地企业安全合规成本全国最低、安全能力水平全国领先，助力粤港澳大 湾区成为全国生成式人工智能安全发展服务最优区域。 南方都市报、南都大数据研究院推出"湾区AI安全发展新引擎"系列报道，深度对话参与联合实验室建设 的专家，一同憧憬大湾区AI安全发展新未来。 "联合实验室将带动大湾区形成完整的AI安全产业集群。"在工业和信息化部电子第五研究所副所长、联 合实验室专家王蕴辉看来，粤港澳大湾区生成式人工智能安全发展联合实验室（简称"联合实验室"）将 发挥"安全基石、协同纽带、创新引擎"三重作用，实现"安全赋能发展"目标，助力生成式AI为大湾区高 质量发展注入新动能。 谈 ...

广州大学网络空间安全学院二级教授、粤语语料库建设与大模型评测广州市重点实验室主任、联合实验 室专家 齐佳音 "联合实验室将集中发力建设大湾区安全语料库，形成全球粤语语料库高地，利好人工智能发展。"广州 大学网络空间安全学院二级教授、粤语语料库建设与大模型评测广州市重点实验室主任、联合实验室专 家齐佳音接受南都访谈时表示，粤港澳大湾区生成式人工智能安全发展联合实验室(简称"联合实验室") 将在"政产学研用"融合攻关下，探索出一套高效、经济、一致的生成式AI安全发展的大湾区模式。 谈优势 以湾区协同机制推动AI治理 南方都市报（以下简称"南都"）：在您看来，联合实验室在大湾区AI产业生态构建中扮演怎样的角色？ 齐佳音：联合实验室成立意义重大。大湾区涉及广东、香港、澳门三地，政府层面的资源调动机制将更 加重要，有了政府层面的主导推动，资源整合的效率将大大提升。另一方面，AI治理涉及多个学科、 多个领域，联合实验室则是一个能聚合粤港澳三地政产学研用优势资源来进行科研攻关的平台，这不仅 有利于攻克生成式AI安全发展中的普遍性问题，更有利于推动粤港澳三地之间的AI治理协同，探索 出"一国两制三法域"背景下的生成式AI安全 ...

Core Viewpoint - The article highlights the growing importance and demand for PSPI (photo-sensitive polyimide) materials in the semiconductor and OLED display industries, with significant investments and advancements being made by both domestic and international companies to enhance production capacity and meet market needs [2][3][4]. Group 1: Company Developments - Jilin Aolai De Optical Materials Co., Ltd. plans to raise up to 299.8 million yuan, with approximately 239.8 million yuan allocated for the construction of a PSPI materials production base, aiming for an annual capacity of 1,000 tons after three years [2]. - Asahi Kasei, a Japanese giant, plans to invest 16 billion yen (approximately 1 billion euros) to double its PIMEL™ PSPI production capacity by 2030 to meet the surging demand from advanced technologies [2]. - Domestic companies such as Wanrun Co., Ltd. and Aolai De have successfully validated their PSPI materials with leading panel manufacturers, indicating a shift towards local production and reduced reliance on imports [4][5]. Group 2: Market Trends and Innovations - PSPI is increasingly recognized as a critical material in high-end manufacturing sectors, including semiconductor packaging and flexible displays, with major suppliers being Asahi Kasei, Toray, and JSR [3]. - Recent innovations include Toray's STF-2000 PSPI, which allows for high-resolution patterning on films up to 200 micrometers thick, targeting new MEMS device applications [3]. - Companies like Ba Yi Space and Dinglong Co., Ltd. are advancing their PSPI products, with Ba Yi Space entering trial production and Dinglong Co., Ltd. achieving significant client certifications for their PSPI offerings [5][6]. Group 3: Investment and Future Outlook - Ba Yi Space is expected to generate orders in the second half of the year as it completes its PSPI production line, while Dinglong Co., Ltd. is focusing on expanding its patent database and product offerings in the semiconductor materials sector [5][6]. - Aisen Co., Ltd. has secured its first wafer factory order for its positive PSPI, with plans to achieve a production capacity of 2,000 tons, potentially generating over 3 billion yuan in revenue [8]. - The domestic PSPI industry is accelerating commercialization, with multiple companies reporting progress from production to client validation, indicating a robust growth trajectory in the sector [4][5].

Core Insights - The overall performance of Sichuan A-share listed companies in the first half of 2025 shows a total revenue of 549.6 billion yuan, a year-on-year increase of 2.61%, and a net profit attributable to shareholders of 56.4 billion yuan, with a growth rate of 7.02% [2] Group 1: Traditional Industries - The food and beverage sector, particularly the liquor industry, has shown robust performance, with Wuliangye leading with a net profit of 19.49 billion yuan and a sales gross margin of 99.52% [2] - Luzhou Laojiao follows with a net profit of 7.66 billion yuan and a sales gross margin of 87.09% [2] - Shede Distillery reported a net profit of 443 million yuan with a sales gross margin of 65.71% [2] Group 2: Emerging Sectors - New energy sectors faced challenges, with Tongwei Co. reporting a net loss of 4.96 billion yuan due to a 42% drop in photovoltaic silicon material prices [2] - Emerging sectors like military, communication, and hydrogen energy have shown rapid growth, with Shudao Equipment's net profit soaring by 5972.30% to 10.15 million yuan [2] - NewEase's net profit reached 3.94 billion yuan, a year-on-year increase of 355.68%, driven by a 350% increase in shipments of high-speed optical modules [2] Group 3: R&D Investment - Total R&D expenses for the 179 Sichuan companies reached 12.35 billion yuan, a year-on-year increase of 12.3%, significantly outpacing revenue growth [2] - Tongwei invested 533 million yuan in photovoltaic technology, while Dongfang Electric increased R&D spending by 16.45% focusing on hydrogen energy projects [2] Group 4: Performance Disparity - There is a notable performance disparity among Sichuan companies, with leading firms like Sichuan Changhong achieving a revenue of 56.7 billion yuan and a net profit increase of 78.6% [2] - Conversely, companies like *ST Lihang and *ST Zhisheng reported losses of 44.57 million yuan and 32.88 million yuan, respectively, highlighting the challenges faced by some firms [2] - 41% of the 179 companies experienced a net profit decline of over 10% year-on-year, indicating significant pressure from macroeconomic conditions and market competition [2][3]

Core Insights - The virtual digital human industry has shifted from initial hype to facing significant operational challenges, including high costs and low returns, leading to a decline in market interest [1][2][3] Group 1: Industry Trends - The rise of virtual beauty influencers like "Liu Yexi" in 2021 sparked a wave of brand engagement with virtual endorsers, leading to a surge in stock prices for related companies [2] - The initial belief in quick returns from virtual digital humans has been challenged by high production costs and diminishing user interest, resulting in many virtual endorsers being removed from platforms [2][3] - A report from QuestMobile indicates that in 2023, the GMV of virtual streamers was less than one-fifth that of real streamers, with a significant drop in average viewing time and a high fan attrition rate [3] Group 2: Technological Advancements - The development of generative AI has led to the evolution of virtual digital humans into "smart humans," utilizing advanced technologies for more human-like interactions [4] - Companies are leveraging modular tools and SaaS platforms to reduce production costs and deploy digital humans in practical applications across various sectors, moving away from purely entertainment-focused roles [4][5] Group 3: Market Expansion - Despite existing challenges, the virtual digital human market is projected to grow significantly, with estimates suggesting a core market size exceeding 48 billion yuan by 2025 [6] - Investment activity in the sector has increased, with 23 funding cases reported in 2025, totaling over 3.5 billion yuan, indicating renewed interest from capital markets [6] - Government initiatives are supporting the development of the digital human sector, with various regions launching new digital human projects to enhance service delivery [6] Group 4: Legal and Ethical Considerations - The emergence of legal issues surrounding virtual digital humans, including copyright and data privacy concerns, is becoming more prominent, as evidenced by recent court rulings [7] - Platforms are enhancing governance measures to mitigate risks associated with AI-generated content, including the identification and removal of misleading accounts [7] Group 5: Future Outlook - The consensus in the industry is that within the next five years, virtual digital humans will transition from being seen as mere novelties to becoming essential tools for digital transformation and economic growth [7]

Core Viewpoint - The first robot PhD student in China, "Xueba 01," has officially enrolled at Shanghai Theatre Academy, showcasing advancements in robotics and AI integration in the arts [6][8]. Group 1: Enrollment and Interaction - "Xueba 01" successfully completed the enrollment process and received a student ID, engaging in interactions with faculty and students [4][6]. - The robot demonstrated quick responses and interactive capabilities, surprising attendees with its ability to engage in conversation [6][5]. Group 2: Background and Development - "Xueba 01" is an optimized version of the "Xingzhe No. 2" robot, which previously won third place in a half-marathon for humanoid robots [6]. - The robot's design includes realistic facial expressions and skin made from silicone, allowing for a lifelike appearance [6]. Group 3: Educational Framework - The educational program for "Xueba 01" includes modules on perception, language generation, artistic expression, and ethical modeling, integrating generative AI into theatrical arts [8][9]. - The robot will undergo a four-year curriculum to learn basic actions, performance techniques, and develop its own artistic works [9][11]. Group 4: Future Aspirations - The goal is for "Xueba 01" to reach a professional level in specific artistic fields, potentially surpassing human actors in certain areas [11]. - The robot is expected to create a representative work by the end of its studies, aiming to impress audiences with its performance [11][13].