责任编辑：何俊熹 对于是否年内发布AI智能体一事，截至发稿，DeepSeek方面暂无回应。 新浪科技讯 9月4日下午消息，据外媒报道，中国人工智能公司DeepSeek计划在2025年底前正式发布AI 智能体（AI Agent），以抢占下一代人机交互入口。 根据公开资料整理，AI智能体被视为继大模型之后的重点赛道，具备自主决策、任务拆解与跨应用协 同能力。 ...

根据公开资料整理，AI智能体被视为继大模型之后的重点赛道，具备自主决策、任务拆解与跨应用协 同能力。 责任编辑：李昂 9月4日，有消息称，中国人工智能公司DeepSeek计划在2025年底前正式发布AI智能体（AI Agent），以 抢占下一代人机交互入口。 新浪科技就此询问DeepSeek，公司暂无回应。 ...

新华财经北京9月4日电（记者吴丛司）《关于深入实施"人工智能+"行动的意见》（以下简称《意 见》）日前正式发布，明确提出推动软件信息服务企业智能化转型，重构产品形态和服务模式，培育人 工智能应用服务商，发展"模型即服务""智能体即服务"等新模式，打造人工智能应用服务链。 宇信科技首席技术官张宁日前接受新华财经专访时表示，AI智能体（AI Agent）作为一种新兴的技术范 式，正在金融科技领域展现出成为下一代核心竞争力的潜力。银行业这一始终处于技术演进前沿的领 域，面临着客户需求日益多样化、运营成本持续上升，行业竞争不断加剧等挑战，更需要引入高效的技 术手段以提升服务能力和运营效率。 在信贷业务中，AI Agent可自动化处理大量无序扫描文件，完成自动分类、排版与信息提取。张宁介 绍，该系统能够识别文件类型、整理多页合同，提取如企业名称、收入等关键信息，并实现信息溯源， 明确数据来源，保障数据的准确性与可信度。 在数据业务方面，宇信科技通过自研的ChatBI平台，不仅将自然语言转化为分析语言，还能主动为管理 者推送数据，将业务问题转化为可分析的知识结构。该平台能够预判分析需求，主动推荐后续问题，帮 助业务人员降 ...

消息面上，8月11日，出门问问发布公告，预计截至2025年6月30日止6个月(本期间)，期内亏损较去年 同期的人民币5.79亿元可能实现不少于99%的同比减少，主要原因为或然可赎回优先股及普通股的账面 值变动。经调整净亏损(非国际财务报告准则计量)较去年同期经调整净亏损的人民币5590万元可能实现 不少于96%的同比减少，接近盈亏平衡。 根据目前可得资料，董事会认为本期间亏损变动的主要原因为：本集团积极推动"组织AI化"的转型，将 AI智能体(AI Agent)深度融入核心业务流程与运营管理体系，构建了高效、灵活、自我进化的"AI原生工 作流"。同时通过持续优化资源配置、积极管控各项费用等措施有效提升运营效率。 出门问问(02438)高开逾5%，截至发稿，涨5.49%，报0.96港元，成交额69.6万港元。 ...

Core Viewpoint - The company, Outermost Inquiry (02438.HK), anticipates a significant reduction in losses for the six months ending June 30, 2025, with a potential decrease of at least 99% compared to the same period last year, primarily due to changes in the book value of redeemable preferred shares and common stock [1] Financial Performance - The expected adjusted net loss (non-IFRS measure) is projected to decrease by at least 96% compared to the adjusted net loss of RMB 55.9 million from the same period last year, approaching breakeven [1] Strategic Initiatives - The board attributes the change in losses to the company's active push towards "organizational AI transformation," integrating AI agents deeply into core business processes and operational management systems [1] - The company has established an efficient, flexible, and self-evolving "AI-native workflow" as part of its transformation strategy [1] - Continuous optimization of resource allocation and proactive cost management measures have effectively enhanced operational efficiency [1]

Group 1 - The AI sector is experiencing several trends this year, including the transition of large models from research to industry applications, and the shift from basic algorithmic computing power to practical applications [3] - Key application areas expected to see significant growth include AI Coding, AI Agents, and AI Hardware, which can create real commercial value and profitability [3] - The pharmaceutical industry is increasingly recognizing AI as a crucial asset for drug development, with algorithms and AI-generated experimental data becoming more important than traditional assets like patient data [3] Group 2 - The healthcare sector is highlighted as a major area of investment, with significant breakthroughs anticipated in cancer treatment over the next 3 to 5 years [3] - Current valuations of AI startups have returned to a more rational state compared to previous years, with a narrowing focus on top companies that can secure larger funding amounts [3] - The investment landscape has shifted from a broader pool of companies to a concentrated focus on fewer firms, with only about 10 companies now attracting significant investment compared to around 50 previously [3]

Core Insights - The article discusses the rising prominence of AI agents, predicting that 2025 will be the "Year of AI Agents" with significant market growth expected [1][3] - The AI agent market is projected to grow from $5.1 billion in 2024 to $47.1 billion by 2030, reflecting a compound annual growth rate (CAGR) of 44.8% [1] - The launch of Manus, a general-purpose AI agent, has sparked interest and investment in the AI agent sector [5] Market Dynamics - AI agents are defined as intelligent entities capable of autonomous decision-making, environmental perception, task execution, and learning [3] - The core technological breakthroughs in AI agents include cross-domain task decomposition, multi-modal data fusion, and cloud-based heterogeneous computing frameworks [3][4] - Major tech companies are actively developing AI agents, with OpenAI planning to offer premium AI agent services tailored for high-end enterprise needs [6] Challenges in Deployment - Companies face challenges in integrating AI agents with existing systems, ensuring data security, and gaining user acceptance [6] - The deployment of AI agents requires a strategic overhaul of management models and business processes to leverage advanced AI technologies effectively [6] Technological Evolution - Large models play a crucial role in supporting AI agents, providing capabilities in language understanding, logical reasoning, and knowledge storage [4] - The evolution of AI is moving from tools to digital assistants, with AI agents becoming more capable of real-time decision-making and interaction with the physical world [4] Value Creation - AI agents are penetrating various industries, driven by advancements in AI technology from discriminative to generative and reasoning AI [8] - Companies are encouraged to focus on high-value scenarios and solid data foundations to accelerate the deployment of AI agents [8] - The integration of large models, AI agents, and tool ecosystems is essential for creating complex interactive networks that enhance business operations [8][9]