The Briefing - Enterprise Agents
文章语言：
简
繁
EN
Share
Minutes
原文
会议摘要
Industry leaders discuss AI's impact on sectors, emphasizing its role in enhancing efficiency, productivity, and knowledge work. By 2025, AI agents like Claude revolutionize software development, reducing time-to-market. Enterprises are urged to integrate AI, adapt, and customize tools for optimal outcomes, as exemplified by companies like Spotify and Salesforce.
会议速览
Revolutionizing Enterprise Work: AI Agents' Journey from Hype to RealityIn 2025, AI agents like Claude transformed software development by writing significant portions of code, proving their real-world value. Building on this success, 2026 sees similar advancements in knowledge work, amplifying productivity across all enterprise functions. By enabling real insights and focusing on critical tasks, these technologies are redefining professional efficiency and institutional expertise.
Revolutionizing Business with Claude: Smarter Employees, Faster Processes, and Transformative ProductsClaude, a thinking engine, enhances business operations by integrating into everyday tools for smarter employees, optimizing processes across operations for faster outcomes, and enabling new product capabilities for customer value and revenue growth. This holistic approach ensures sustained competitive advantage, contrasting with piecemeal AI solutions.
Revolutionizing Enterprises with AI: Spotify, Novo Nordisk, and Salesforce Leading the TransformationThe dialogue highlights how leading companies are leveraging AI to enhance efficiency and innovation. Spotify is automating code migrations, Novo Nordisk is accelerating regulatory documentation, and Salesforce is improving customer interaction. These advancements are driven by partnerships and AI tools, showcasing the transformative power of AI in enterprise operations.
Expanding Claude's Capabilities for Enterprise Knowledge Workers with CoworkClaude's recent updates introduce Cowork, enabling enterprise knowledge workers to delegate complex tasks, ensuring polished, near-final deliverables. This advancement integrates organizational standards, quality, and work methods, enhancing project completion and team efficiency.
Revolutionizing Enterprise Workflows: Introducing Customizable Plugins and Enhanced Tools for Cowork and ClaudeIntroduces customizable plugins for Cowork and Claude, enhancing enterprise readiness with admin controls, private marketplaces, and streamlined user interactions. New connectors and plugins expand capabilities, enabling seamless integration across various enterprise tools and improving productivity.
Revolutionizing Enterprise Operations: How AI Enhances Efficiency and CollaborationSilver and Capital leverages AI, specifically Claude, to streamline operations, enhance collaboration, and accelerate growth. The finance team identifies competitive blockers, the legal team updates processes with AI assistance, and sales and product teams act swiftly on insights. This integration of AI across departments removes toil, enabling teams to focus on strategy and collaboration, demonstrating the transformative power of AI in enterprise settings.
Leaders in AI Across Finance, Law, Software, and Media Unite for a Fiduciary Grade AI DiscussionA group of AI leaders from various industries, including finance, law, software, and media, gathered to discuss the evolution of AI, particularly focusing on the development of professional and fiduciary grade AI solutions. The conversation highlighted the importance of AI in enhancing services for knowledge workers and fiduciary professionals, with a specific mention of a product developed for this purpose.
AI's Pivotal Role in Reshaping Future Workplaces and IndustriesA discussion unfolds on the transformative impact of AI by 2026, highlighting its critical role in enhancing system resilience and determinism across diverse sectors, from finance and healthcare to technology and exchanges. Participants share insights on integrating AI for strategic advancements, emphasizing its potential to redefine operational paradigms and drive innovation in work environments.
AI's Evolution in Engineering: From Code Completion to Collaborative Agents and Regulatory WorkflowsThe dialogue explores AI's transformative role in accelerating codebase management and regulatory tasks, highlighting the shift from experimental to scaled AI applications. It discusses the evolution of AI from simple code completion tools to collaborative agents capable of handling complex tasks, such as writing tests, refactoring code, and processing large documents. The conversation also touches on the strategic importance of leadership in managing AI projects and the need for a playbook that adapts to the changing dynamics of human-AI collaboration.
Navigating AI Integration: Change Management, Accountability, and Risk Calibration in Fiduciary ProfessionsDiscusses the shift in accountability with AI, the emergence of assembly as a solution paradigm, moving from risk avoidance to calibration, and the importance of leadership in embracing technological change for effective integration in fiduciary sectors.
Codifying Expertise and Overcoming Challenges in Enterprise AI RolloutsThe dialogue emphasizes the importance of leveraging ground-level expertise in AI integration, highlighting a shift from assistive AI applications to more advanced uses like code exploration and automated support ticket generation. It discusses challenges and strategies in codifying domain-specific expertise into scalable AI systems, particularly in capital markets, showcasing examples from market execution improvements and customer-facing AI applications.
Importance of Validation, Verification, and Data Privacy in AI Tool Integration for Business GrowthThe dialogue emphasizes the significance of validating and verifying AI tool outputs for accuracy and the critical need to protect both company and customer intellectual property. These lessons are applied across various departments to ensure internal automation processes in legal, finance, and customer support sectors are secure and reliable, promoting business expansion and maintaining competitive edge.
Building Trust in AI Systems: Organizational Strategies and User ExperienceThe dialogue explores the importance of trust in AI systems, emphasizing the need for verification, ongoing monitoring, and user-friendly design. It highlights the integration of observability, auditability, and human judgment to ensure reliability and confidence in AI applications, particularly in regulated markets.
AI's Transformative Impact on Workplace Efficiency and Healthcare QualityDiscusses AI tools like Claude and Cowork enhancing non-technical roles, improving healthcare through research integration, and boosting strategic thinking across organizations.
Navigating AI Adoption and Change Management in BusinessThe dialogue emphasizes the importance of fiduciary-grade AI, rapid tool development, and internal change management for business leaders. It advises accelerating AI adoption through strategic planning and customer engagement, highlighting the need for effective change management to fully realize AI's benefits.
Strategies for Effective AI Adoption: Balancing Innovation, Speed, and IP ProtectionEmphasizing personal involvement and investment in AI tools, maintaining a rapid pace of adaptation to technological changes, and safeguarding institutional intellectual property are crucial for successful AI integration. Leaders should foster open exploration to identify opportunities and bottlenecks, encourage experimentation, and ensure that the organization remains competitive while protecting its core assets.
Economic Implications of Emerging Technology: A Deep DiveA segment featuring an expert discussing global economic shifts driven by rapid technological advancements, highlighting insights from leading industry professionals.
Anthropic's Economic Index: Measuring AI's Impact on Global Workforce and ProductivityThe Anthropic Economic Index tracks AI usage, revealing trends in automation, workflow augmentation, and productivity gains. Analyzing Claude's global application, it assesses labor market implications and guides reporting adjustments based on evolving data.
General Purpose Technology's Dual Impact: Automation vs. Augmentation in the WorkforceEconomists highlight large language models' broad economic impact, noting automation's prevalence over augmentation. As businesses integrate these tools, automation dominates, reshaping workflows and labor dynamics, though augmentation offers collaborative potential. This shift underscores the dual nature of transformative technologies, influencing how work is conducted and valued across various sectors.
AI's Uneven Impact on Knowledge Work: Productivity Gains and Displacement RisksAI is reshaping jobs, offering significant productivity boosts to high-skilled roles while posing displacement risks to more routine tasks. The conversation highlights the need for monitoring labor market changes and adapting strategies to amplify rather than replace human work, especially in complex, knowledge-intensive environments.
Adapting Roles and Leveraging AI for Enhanced Question Asking and PrototypingThe dialogue explores how AI, specifically Claude, facilitates rapid prototyping and question asking, transforming roles such as economists and product managers. It highlights the importance of adaptability, curiosity, and oversight in a dynamic work environment where traditional job boundaries are evolving.
Leveraging AI for Enhanced Enterprise Performance: The Role of Context and Organizational AdaptabilityThe dialogue highlights the importance of context and organizational adaptability in leveraging AI, such as Claude, for complex tasks. It emphasizes the need for modern data ecosystems and organizational processes to effectively integrate AI capabilities, suggesting that adaptable organizations will benefit significantly.
Navigating Rapid Technological Change: Embracing Uncertainty and AdaptabilityDiscusses the transformative impact of a general purpose technology, emphasizing its fast adoption and the necessity for humility and adaptability in a rapidly evolving environment. Highlights the potential for innovation in innovation itself, urging preparedness for uncertainty in both the near term and long run.
AI's Future in Knowledge Work: Anthropic's Vision and Practical Applications with ClaudeA forward-looking dialogue highlights AI's transformative role in knowledge work by 2026, featuring discussions on Claude's potential as a thinking engine. It previews upcoming sessions detailing Claude's internal usage across various departments and its deployment strategies in enterprise settings, emphasizing live demonstrations and expert insights.
要点回答
Q：What was the killer use case for AI in 2025 as indicated by the speech?
A：The killer use case for AI in 2025 was in software development, with AI agents significantly transforming coding processes.
Q：How did AI transform software development and what were the results?
A：AI transformed software development by enabling AI to write 90-100% of the code, with enterprises shipping in weeks instead of many quarters. This resulted in real value add beyond incremental gains and led to the success of fast-growing startups powered by cloud technology.
Q：How is cloud expected to transform knowledge work in 2026?
A：Cloud is expected to transform knowledge work in 2026 by bringing the same power to knowledge workers as Claude Code did to software developers, enabling them to delegate hard challenges and focus on work that truly matters.
Q：What is the role of Claude in enterprise transformation?
A：Claude is the thinking engine that powers the transformation in enterprises, designed to give smarter employees, faster processes, and transformative products by integrating into tools used by knowledge workers.
Q：How is Claude being embedded in the tools used by knowledge workers?
A：Claude is embedded in tools such as Excel, PowerPoint, Slack, and financial models, allowing for faster processes and enabling entire teams, not just specialists, to perform tasks that previously took hours.
Q：What advantages do companies gain by integrating Claude into their operations?
A：Companies gain the advantage of transforming regulatory submissions, clinical reports, and customer support from operational bottlenecks into massive company advantages by integrating Claude into their operations, which also helps in creating transformative products for customers.
Q：What is the thinking divide, and how are enterprises approaching AI?
A：The thinking divide refers to the difference between enterprises that move beyond experimentation and treat AI as core to their operations and those that only implement AI as a point solution for a single workflow. The former is leading the pack in innovation.
Q：Which companies are leading the transformation, and what are their specific use cases?
A：Spotify, Novo Nordisk, and Salesforce are leading the transformation. Spotify's engineers use Claude to modernize code across services with up to a 90% reduction in engineering time. Novo Nordisk uses an AI-powered platform, Novo Scribe, to produce regulatory-grade content, with documentation creation going from 10+ weeks to 10 minutes. Salesforce leverages Claude models to enhance AI in Slack for customer data navigation with a 96% satisfaction rate for their tools.
Q：What is the purpose of the Claude Enterprise product?
A：The purpose of the Claude Enterprise product is to help knowledge workers work smarter by providing them with AI capabilities to make the most of AI across their enterprise.
Q：What are the capabilities that have been expanded for Claude?
A：The capabilities expanded for Claude include delivering polished near-final work, suggesting drafts, and assisting with actual completed projects and deliverables.
Q：What are the new updates for Cowork and plugins?
A：The new updates for Cowork and plugins include the ability to deliver polished near-final work, share specialized skills, context, and connections to other enterprise tools, and the introduction of controls to customize plugins for an organization's specific needs.
Q：How are enterprises customizing Claude?
A：Enterprises are customizing Claude through plugins, which enable them to align with their company standards, quality bars, and ways of working.
Q：What controls are provided for administrators to build and customize plugins?
A：Administrators can build customized plugins by using starter templates or creating plugins from scratch. They can set up connectors to streamline the process and customize skills, commands, and Mcps within plugins.
Q：How can users interact with plugins?
A：Users can interact with plugins using slash commands, structured workflows, and structured forms, making the process intuitive and easy.
Q：What changes have been made to the MCP experience and connectors?
A：The MCP experience has been overhauled to improve discoverability, streamline admin controls, and enhance the core. This includes new connectors from various enterprise software companies and expanded plugin templates.
Q：What is the capability of Claude in editing files and passing context?
A：Claude can now edit files and pass context between various apps like cowork, Excel, and PowerPoint, even across multiple files in the same application, without starting over when switching between apps.
Q：What was the example given of how Silver and Capital used Claude?
A：The example given is that the finance team at Silver and Capital used Claude to identify growth blockers and surface actionable insights across customer touchpoints. The legal team used a Claude-powered agent to update processes and gain insights from previous agreements, which were then operationalized across teams for faster action.
Q：What is the role of AI in the professional services according to Steve Asci from Thompson Roys?
A：Steve Asci from Thompson Roys sees AI as a transformative tool in professional services, specifically in the fiduciary grade AI business, which they use to develop products like Cosco Imci. He mentions that AI is being integrated into their engineering processes, writing tests, refactoring legacy code, and creating internal AI agents that can execute tasks from start to finish.
Q：How is AI being utilized in trading platforms according to the NYSE's representative?
A：The NYSE's representative explains that AI is being used to process large amounts of information, such as in regulatory and market watch workflows. Specific applications include reviewing proxy filings, auditing SCC filings, and news classification. The NYSE has announced an initiative platform for tokenized trading, where AI is part of the process to accelerate implementation.
Q：What are the observed changes in AI usage over the past 18 months?
A：Over the past 18 months, AI has shifted from being primarily a chat interface used for code completion to a more independent and collaborative tool with agency and reasoning capabilities, as noted by the NYSE's representative.
Q：What challenges and opportunities does AI present for leaders in finance and technology?
A：Leaders in finance and technology are facing challenges such as reworking processes to take advantage of AI tools, shifting accountability towards continuous monitoring of AI behavior and outcomes, and managing an assembly of multiple models and vendors. Opportunities include the potential to calibrate risk in AI systems, the shift towards probabilistic accountability, and the need to build and calibrate risk systems differently from traditional risk avoidance.
Q：How are enterprise organizations planning to incorporate and utilize AI within their existing systems and workflows?
A：Enterprise organizations are looking to codify subject matter expertise into AI systems and organizational workflows. They are exploring ways for ground-level staff to identify and utilize information to enhance customer service, and leveraging AI tools such as the Claude 4 and Claude code harness for more advanced code exploration. Additionally, organizations like the NYSE are planning to implement AI into market structures, trading platforms, and various business operations such as mortgage businesses and fixed income data.
Q：What are the key learnings from the co counsel rollout that are being applied internally?
A：The key learnings from the co counsel rollout include the importance of validation and verification of the tool's output, ensuring the accuracy of the AI's results, and the implementation of internal safeguards to prevent intellectual property from being shared with competitors or others. These learnings are being applied to internal automation in legal, finance, support, customer success, engineering, and other departments.
Q：How does the speaker's organization ensure trust in AI systems and the protection of intellectual property?
A：The speaker's organization ensures trust in AI systems by adhering to core trust principles, such as unbiased, fact-based information, and by modeling agnosticism, selecting the best enterprise models to support their efforts. They are cautious about the use of AI to protect intellectual property and customer data, not allowing it to leave their organization's control.
Q：What is the role of trust in the products mentioned by the speaker?
A：Trust is a foundational element in the products described by the speaker. The products are grounded in unbiased, factual content and domain expertise, and they embody the trust principles established by Reuters, which the speaker's organization has adopted. This trustworthiness is paramount in the development of their offerings.
Q：What are the considerations mentioned regarding the user experience and building trust with AI capabilities?
A：When considering AI capabilities, there's an emphasis on ensuring the user experience by building trust, particularly in production settings. This involves verifying capabilities before implementation and continuously monitoring their use. The goal is to create a reliable user experience that allows both organizations and end-users to have confidence in the AI's functionality.
Q：How is observability, auditability, and explainability integrated into the use of AI in regulated markets?
A：In regulated markets, the speaker's organization is focusing on bringing observability, auditability, and explainability to AI usage. This approach helps maintain compliance and ensures that human judgment and decision-making complement the AI's recommendations.
Q：What transformations have been observed with the use of Claude and cowork within the organization?
A：Transformations observed with the use of Claude and cowork include enhanced capabilities in development teams creating and upgrading products, and non-technical staff using the technology in various ways not initially anticipated. The adoption of cowork plugins is shifting the role of agents from assistants to collaborators, and the application of Claude and cowork to general office functions like finance and internal audit is improving speed, accuracy, and reliability.
Q：What are the expected benefits of AI in healthcare according to the speaker?
A：The speaker believes that AI will improve the quality of care and expand the healthcare workforce, particularly to care for an aging population with various conditions. These improvements and expansions in access are expected to become apparent by 2026.
Q：What is the anticipated role of AI in financial services?
A：The anticipated role of AI in financial services includes its use as a 'fiduciary-grade' tool, meaning it will become a necessity for lawyers and accountants to get AI right to avoid costly errors, with rapid tool development expected.
Q：What advice does the speaker offer to business leaders looking to accelerate their AI adoption?
A：The speaker advises business leaders to focus on open exploration within their organization to discover opportunities and bottlenecks in their processes. They should also keep pace with AI innovation through experimentation and reimagining existing solutions, emphasizing personal involvement and investment in using AI tools, moving quickly, and protecting intellectual property.
Q：What is the purpose of the Anthropic Economic Index?
A：The purpose of the Anthropic Economic Index is to understand how AI is being used across the economy. It is a comprehensive data set that uses privacy-preserving methods to analyze how people and businesses are using AI in personal and professional settings, offering insights into automation, workflow augmentation, efficiency gains, and productivity benefits.
Q：How has the scope of AI's impact on the economy changed over the past year?
A：Over the past year, the scope of AI's impact on the economy has broadened significantly. About two-thirds of all jobs in the U.S. economy now have at least a quarter of their tasks being performed by AI, up from roughly a third a year ago. This indicates that the tools and technology are becoming more capable, and businesses and individuals are experimenting and adopting these tools for new purposes.
Q：What distinguishes automated use of AI from augmented use, and how might this affect the economy?
A：Automated use of AI involves a one-way task where the AI completes the job without further interaction, like translating a document. Augmented use, on the other hand, involves a more collaborative workflow where the AI provides feedback and assistance in a back-and-forth interaction, such as evaluating logic and structure of an argument. The nature of tasks becoming automated or augmented will shape the AI's impact on the economy and labor market.
Q：How are businesses embedding CLA in their workflows through the API?
A：Businesses are embedding CLA in their workflows through the API in automated ways, similar to how transformative technologies typically spread throughout the economy.
Q：What does the data indicate about AI's impact on knowledge work and enterprise knowledge workers?
A：The data suggests that AI is changing knowledge work and there is a risk of displacement for enterprise knowledge workers, particularly those从事纯实施性工作或使用AI进行任务的岗位如数据录入员和技术撰稿人。
Q：What is the labor market implication of AI according to recent work and reports?
A：Recent work and reports indicate that the labor market implications of AI are likely to be very uneven, with uneven impacts similar to past waves of IT innovation. High-skilled workers with more years of schooling may see larger productivity gains, but there's a concern about jobs that involve pure implementation.
Q：What tasks are AI systems like Claude particularly good at, and which jobs might be at higher risk of displacement?
A：AI systems like Claude are particularly good at tasks involving reading unstructured text and combining, extracting, and structuring information. Jobs that involve implementation tasks, such as data entry workers and technical writers, might be at a higher risk of displacement.
Q：Is there any evidence of widespread displacement in the labor market?
A：There is currently no evidence of widespread displacement in the labor market, as the unemployment rate in the US is near what many economists consider full employment.
Q：What does the use of AI in higher skilled work mean for professionals trying to enhance their work?
A：Professionals can leverage AI for enhanced productivity by focusing on high-leverage tasks such as setting direction, asking insightful questions, and iterating quickly. This shift allows professionals to extend their capabilities with technology like Claude and adapt their roles to include new elements such as creating interactive dashboards.
Q：What changes to job roles and skills are anticipated in the new AI-driven environment?
A：The new AI-driven environment is expected to bring changes to job roles and skills, with an increased emphasis on adaptability, curiosity, and the ability to delegate and oversee complex tasks as models become more sophisticated.
Q：What does the data indicate about enterprises' use of AI and their data ecosystems?
A：The data indicates that enterprises using AI for complex tasks rely on contextual information, suggesting that having the right sort of data ecosystem is crucial. Modernizing data infrastructure and possibly organizational processes and structures may be necessary to provide the necessary information to AI models.
Q：What advice is given to leaders of enterprise organizations based on the data and research?
A：Leaders of enterprise organizations should focus on embedding AI capabilities effectively within their organizations, which may require complementary investments in data modernization efforts and organizational changes to centralize and utilize tacit information.
Q：What are the broader implications of AI adoption across the economy?
A：The broader implications of AI adoption include a transformation across the economy with fast-paced automation, discovery of new ways to do things, and immense uncertainty regarding the near and long-term future.
