王少康分享组织级AI的底层逻辑：身份、权限、标准是前提

Overview: AI-powered office software has entered the organizational-level agent stage. WPS Comate has evolved from a point tool into a digital employee, with identity recognition, permission inheritance, security controls, and process auditing as prerequisites. For ordinary users, this means office software is no longer just for typing and layout, but an AI assistant that can enter core workflows and execute tasks on their behalf.

1. Efficiency gains: Product quality inspection reports can now be generated in 10 minutes instead of 2 days; talent reviews can be viewed daily instead of once a year; and state-owned asset supervision data for one province moved from annual aggregation to weekly updates. The Zhongshan government platform used 600,000 private-domain documents to generate official documents, reducing drafting time from 3 hours to 40 minutes and cutting formatting errors by 75%.

2. Technical foundation: Enterprise context is broken down into six categories – knowledge, data, time, relationships, rules/processes, and boundaries – corresponding to data bottlenecks, experience bottlenecks, and action bottlenecks. AI must combine these contextual elements to execute tasks safely.

3. Market signal: In Q1 2026, WPS's share of China's domestic office suite market reached 32.1%, surpassing Microsoft's 28.5%. WPS AI also ranked first in spreadsheet-processing benchmarks, showing that AI capability has become the real competitive focus in office software.

Overview: Organizational-level AI office software is becoming a new competitive track for office software brands. Brand teams should watch product positioning, R&D direction, shifting user needs, and market-share signals.

1. Positioning has moved from component-level AI and knowledge-base AI to agent-based AI. WPS Comate is positioned as an organizational digital employee that enters core office workflows, suggesting brands should differentiate around an "executor" image rather than offering mere features.

2. Product development should emphasize the six-way decomposition of enterprise context to resolve data, experience, and action bottlenecks. By supporting private deployment and multi-model routing to fit Chinese users, brands can design localization capabilities accordingly.

3. User-behavior observations show government and enterprise clients value security, standards, evaluation, and business connectivity. WPS has strengthened trust by passing the first round of CAICT evaluations and participating in standards development; brands can adopt a "standards-first" endorsement strategy.

4. In market share, WPS surpassed Microsoft with 32.1% versus 28.5%, and WPS AI leads in spreadsheet benchmarks. AI capability has become the key battleground, so brands need to accelerate their agent-product roadmaps.

Overview: Organizational-level AI office software creates both new sales opportunities and compliance requirements. Sellers should track the growth of the office-agent market, changes in customer demand, partnership entry points, and potential risks.

1. Growing market: In Q1 2026, WPS overtook Microsoft in China's domestic office-suite market share. Office suites are becoming the core entry point for agents in daily work, driving demand for organizational-level AI deployments.

2. Changing customer demand: Enterprise clients want AI to solve data, experience, and action bottlenecks. Cases include quality inspection reports cut from 2 days to 10 minutes, talent reviews available daily, and state-asset data updated weekly. Sellers can recommend solutions targeting these scenarios.

3. Partnerships and support: WPS Comate and WPS 365 demonstrate organizational-level AI offerings that can serve customers through private deployment and multi-model routing. WPS has also participated in CAICT evaluations and standards work; sellers can use "passed standard evaluation" as a sales point.

4. Risk reminder: AI deployment requires identity recognition, permission inheritance, security controls, and process auditing. Sellers need to help customers meet security and compliance requirements; otherwise, AI cannot enter core workflows.

Overview: Factories can look at how organizational-level AI is deployed in production management scenarios and what supporting conditions are needed for digitalization.

1. Product quality inspection reports were cut from 2 days to 10 minutes, showing that AI can sharply reduce quality-document processing time. Factories can evaluate integrating an agent into their inspection workflow.

2. Talent reviews moved from once a year to daily visibility, and state-owned asset supervision data moved from annual to weekly updates. This suggests AI enables high-frequency reporting, HR data, and regulatory data, giving management faster decision-making.

3. Digitalization implications: WPS divides enterprise context into six categories – knowledge, data, time, relationships, rules/processes, and boundaries – corresponding to data, experience, and action bottlenecks. Factories need to remove these bottlenecks and ensure identity recognition, permission inheritance, security controls, and process auditing.

4. Business opportunity: The office suite acts as an entry point for agents, with support for private deployment and multi-model routing. Factories can combine WPS AI's leading spreadsheet capabilities to optimize production reports, inspection records, and other operational steps.

Overview: Service providers' clients are moving from point AI tools to organizational-level agents. The industry trend, new technologies, customer pain points, and solution paths are all becoming clear.

1. Industry trend: AI office is shifting from single-feature enhancement to the restructuring of organizational workflows and governance systems. Office suites have become the core entry point for agents in daily work; China's path emphasizes local files, collaboration entries, private deployment, and multi-model routing.

2. New technology: WPS Comate has evolved into an executor, entering core processes as an organizational digital employee. The six-way decomposition of enterprise context (knowledge, data, time, relationships, rules/processes, and boundaries) is a key technical framework.

3. Customer pain points: data bottlenecks, experience bottlenecks, and action bottlenecks. AI deployment requires identity recognition, permission inheritance, security controls, and process auditing, so service providers need to help clients build these foundations.

4. Solutions: WPS 365 demonstrates a complete landing path for organizational-level AI, including standard evaluation (first CAICT evaluation), unified identity identifiers, and participation in standards development. Service providers can use these solutions to deliver deployment and integration services.

Overview: Organizational-level AI places higher demands on platforms. Marketplace operators need to pay attention to customers' requirements for security, standards, evaluation, and business connectivity, as well as the latest platform practices.

1. Platform requirements: AI office deployment is premised on identity recognition, permission inheritance, security controls, and process auditing. Platforms must solve boundary issues in enterprise context before AI can enter core workflows.

2. Latest platform practices: WPS has upgraded Comate into an organizational digital employee, decomposed enterprise context into six categories (knowledge, data, time, relationships, rules/processes, boundaries), and showcased organizational-level AI office solutions through WPS 365.

3. Standards and risk control: CAICT has released an agent capability evaluation system; WPS passed the first batch of evaluations and obtained a unified identity identifier. Platforms should follow standards and evaluations to manage compliance risk, while also paying attention to China-specific paths such as local files, private deployment, and multi-model routing.

4. Ecosystem and partner recruitment: WPS has overtaken Microsoft in China's domestic office-suite market share and leads in spreadsheet AI benchmarks. These signals can help platform operators decide where to focus ecosystem partnerships and give priority to vendors with organizational-level AI capabilities.

Overview: Organizational-level AI office software is creating new industrial dynamics and raising new questions around security, standards, and governance that deserve researchers' attention.

1. New industrial dynamics: AI office is moving from single-feature enhancement to the restructuring of organizational workflows and governance systems. Office suites have become the core entry point for agents in daily work; China's path emphasizes local files, collaboration entries, private deployment, and multi-model routing.

2. New questions: AI entering core workflows must meet prerequisites such as identity recognition, permission inheritance, security controls, and process auditing. Enterprise context needs to be decomposed into six categories – knowledge, data, time, relationships, rules/processes, and boundaries – corresponding to data, experience, and action bottlenecks.

3. Policy, regulation, and standards: China's CAICT has launched an agent capability evaluation system. WPS passed the first batch of enterprise-level office agent evaluations and obtained a unified identity identifier. Relevant parties participated in drafting the "Guide for Building AI Office Large Model Systems" and the "Technical Requirements for AI Office Large Model Systems" under the key AI technologies framework, indicating that standards are being advanced.

4. Business models: WPS surpassed Microsoft with 32.1% market share, and WPS AI leads in spreadsheet benchmarks. Government cases show that AI can greatly improve the efficiency of official document generation. Business models are shifting from selling tools to providing organizational digital employees and services.