豆包、千问将下线智能体功能

This article covers the key update that China’s two leading AI platforms, Doubao and Tongyi Qianwen, are simultaneously phasing out their AI agent functionality. Below is a summary of core information and practical guidance:

1. Key timeline: Both platforms have scheduled the full shutdown for July 15, 2026. For Tongyi Qianwen, anthropomorphic interactive AI agents and user-built agents may be taken offline as early as July 10. After shutdown, users will no longer be able to create new agents or access existing ones. Doubao will dispose of relevant user data in accordance with its privacy policy after October 15, 2026.

2. Backup guidance: Users who need to retain their content should export agent configurations and back up prompt materials and conversation records in advance. Tongyi Qianwen users need to upgrade to version 6.12.9 or higher to export content via the designated path, or save content via copying and screenshots.

The simultaneous shutdown of AI agent functionality by two leading Chinese AI platforms reflects the latest regulatory trends and industry shifts in China’s AI sector, offering key takeaways for brands developing AI-focused businesses:

1. Clear compliance requirements: China’s Interim Measures for the Administration of Artificial Intelligence Anthropomorphic Interactive Services will officially take effect on July 15, imposing strict restrictions on virtual companion and emotionally dependent AI services. Prior to this, the Qinglang special rectification campaign had already taken down more than 14,00 non-compliant AI agents, making compliance a non-negotiable foundation for all AI businesses.

2. Guidance for product R&D: User-built AI agents carry inherent risks, including copyright infringement, inappropriate content, challenges in minor protection, and enormous content moderation pressure. When developing AI-related products, brands must embed compliance requirements early in the development process, and build robust content moderation and minor protection frameworks from the start.

3. Consumer trend insights: User demand for AI agent functionality remains steady. Brands can develop compliant AI agent services aligned with the adjustments made by leading platforms, and launch regulatory-compliant alternative products to meet unmet market demand.

This event clarifies policy directions, risks and opportunities for sellers operating in AI-related fields. Key takeaways are as follows:

1. Clear policy interpretation: Regulatory rules for AI anthropomorphic interactive services are now formally established, with the official management measure taking effect on July 15. Strong restrictions have been imposed on virtual companions, user-built AI agents and related businesses, and large-scale special rectifications have already been implemented to remove a high volume of non-compliant agents, making the compliance red line unambiguous.

2. Clear risk warning: User-built AI agents are prone to violations including unauthorized virtual companions, infringing character setups, inappropriate content, and misleading consumption. Offering such non-compliant services carries significant risk of penalty and removal from the market, so sellers should proactively avoid these non-compliant business lines.

3. New opportunity outlook: Demand for AI agent functionality among users remains genuine, and leading platforms have already launched compliant alternatives. Sellers can partner with compliant platforms to operate and serve compliant AI agent businesses, and capture opportunities in the newly regulated, compliant market by aligning with industry adjustments.

The shutdown of AI agent functionality by two leading AI platforms offers multiple insights for AI product manufacturing and design factories. Key takeaways are as follows:

1. Adjustment to product development requirements: Regulators have put forward clear compliance requirements for anthropomorphic interactive AI and user-built AI agents. When developing and designing AI-related products, factories must prioritize compliance, and integrate minor分级 protection and content moderation functional modules early in the process to meet regulatory standards.

2. Mapping new commercial opportunities: User demand for AI agent functionality remains stable, and leading platforms have launched compliant alternative products. Factories can partner with leading platforms to meet their demand for compliant supporting products, develop regulatory-compliant supporting solutions, and enter the fast-growing compliant AI service track.

3. Insights for digital transformation: AI sector regulation is maturing rapidly. When advancing AI-related digital transformation initiatives, factories need to keep up with the latest regulatory updates, adjust product development and design directions proactively, mitigate compliance risks in advance, and ensure stable long-term business growth.

This event clarifies industry trends, customer pain points and business directions for AI-related service providers. Key takeaways are as follows:

1. Clear industry development trajectory: China’s AI anthropomorphic interactive service industry has entered a phase of compliance-focused development. With formal regulation now in effect, non-compliant user-built AI agent services are being cleared out. The industry will move toward greater standardization and compliance going forward, and non-compliant businesses will no longer have room to operate.

2. Clear customer pain points: Following the shutdown of original AI agent functionality, a large number of users now need data backup and service migration support, while retaining demand for AI agent creation and usage. All these represent addressable customer pain points for service providers.

3. Guidance for business direction adjustment: When offering AI agent-related services, providers need to build robust content moderation systems, implement identity verification and分级 protection for minors, solve the long-standing industry challenge of moderating massive volumes of user-generated content, and launch regulatory-compliant AI agent services to meet market demand.

This business adjustment by two leading Chinese AI platforms clarifies regulatory requirements, operational directions and risk mitigation guidelines for AI-related platform operators. Key takeaways are as follows:

1. Regulatory outlook and risk mitigation: Chinese authorities have introduced dedicated management measures for AI anthropomorphic interactive services, with clear requirements to control non-compliant AI content. Prior to this, the Qinglang special rectification campaign already removed more than 14,000 non-compliant AI agents. Platforms must prioritize mitigating compliance risks stemming from user-built AI agents, including inappropriate content, copyright infringement, consumption inducement, and inadequate minor protection.

2. Operational and governance requirements: Platforms must strictly implement regulatory requirements, proactively conduct internal compliance audits, enforce identity verification for minors, build robust content moderation mechanisms to handle the pressure of moderating massive volumes of user-generated content, and proactively adjust non-compliant product features.

3. Addressing user needs: After taking non-compliant features offline, platforms must provide users with clear paths to export and back up their data, while launching compliant alternative services to meet user demand for AI agent functionality, retain users, and maintain stable platform operations.

The simultaneous shutdown of AI agent functionality by two leading Chinese AI platforms reflects a new phase of development in China’s AI industry, offering high research value for relevant scholars and analysts. Key insights are as follows:

1. New industry dynamics: After the rapid expansion of China’s AI-generated content industry, regulation has caught up quickly, and leading platforms are proactively adjusting their business by removing non-compliant user-built AI agent functionality. The overall industry is shifting from rapid expansion to compliance-focused development, and is currently undergoing restructuring as leading platforms launch compliant alternative products.

2. New industry challenges: The user-built AI agent model has intractable inherent pain points, including massive content moderation pressure, high risk of non-compliant content, difficulties in minor protection, and potential harms from excessive emotional dependence. These represent new research topics that require solutions for the sustainable development of the AI industry.

3. Insights for policy research: China’s AI regulation has been implemented at a very rapid pace, with a complete enforcement pathway in place that combines formal legislation, special rectification campaigns, and corporate compliance guidance. Researchers can focus their work on topics including the impact of regulatory policy on AI industry development, and pathways to compliant growth for AI businesses.