字节跳动与阿里巴巴将禁用拟人化AI自定义智能体，新规即将落地
英文标题：🔶 ByteDance and Alibaba to disable humanlike AI custom agents as new rules loom
字节跳动的豆包和阿里的通义千问两大AI应用宣布将在7月中旬停用用户自定义的拟人化智能体功能，以遵守将于7月15日生效的《人工智能拟人交互服务管理暂行办法》。该办法针对模拟人类性格并提供持续情感交互的AI服务，旨在防范极端思想、隐私泄露与成瘾风险，但明确排除了客服、教育等不涉及持续情感互动的工具。此举标志着中国正快速为高速发展的AI领域构建监管框架，对依赖个性化角色的开发者和平台产生直接影响。
45.1 KB
35.9 KB
45.6 KB
中文
中国两大面向消费者的人工智能应用——字节跳动旗下的豆包和阿里巴巴集团旗下的通义千问——正着手停用自定义智能体功能。针对拟人化AI交互服务的新规即将生效，这是北京方面为快速增长的AI领域建立更广泛监管框架的一部分。
豆包在周五晚间通知用户，由于“产品功能调整”，其智能体功能将于7月15日下线。10月15日之后，豆包的相关数据将依照公司隐私政策处理，用户将无法再在应用内查看或恢复这些数据。
通义千问也在周六上午发布类似通知，称其“拟人化交互智能体和用户创建的智能体功能”将于7月10日停用，而更广泛的“通义千问智能体功能和服务”将在7月15日下线。停服后，用户将无法再访问相关的智能体设置或此前的对话记录。
此前，两款应用都提供了由官方和用户创建的智能体库，这些智能体可以根据特定任务、技能和说话风格进行定制。用户也可以创建自己的智能体，将一个通用聊天机器人变为具有固定人设和语气的指定助手、导师、角色扮演人物或伴侣。
此次下线的时间点，恰逢《人工智能拟人交互服务管理暂行办法》于7月15日正式施行。该办法于4月发布，监管范围覆盖那些“模拟人类人格特质、思维模式和沟通方式，以提供持续情感交互”的AI服务。
办法明确排除了客服机器人、知识问答、职场助手、教育和科研工具，前提是这些工具不涉及持续的情感交互。
该办法指出的风险包括极端思想、隐私泄露、对身心健康的危害，以及依赖或成瘾等问题。
English / 原文
Two of China’s major consumer-facing artificial intelligence apps,
Doubao and
Qwen, are moving to disable customised agent features, as new rules on humanlike AI interaction services are set to take effect, part of Beijing’s push to build a broader regulatory framework for the fast-growing sector.
informed users in a Friday night notice that its agent feature would go offline on July 15 because of “product function adjustments”. After October 15, Doubao’s related data would be handled in accordance with the company’s privacy policy and no longer be viewable or recoverable inside the app.
also issued a similar notice on Saturday morning, saying that its “humanlike interactive agents and user-created agent functions” would be disabled on July 10, while broader “Qwen agent functions and services” would be taken offline on July 15. Users would no longer be able to access related agent settings or previous conversations after the shutdown.
Both apps had offered a pool of agents, created by both the companies and users, that could be customised for specific tasks, skills and speaking styles. Users could also create their own agents, turning a general-purpose chatbot into a named assistant, tutor, role-playing character or companion with a fixed persona and tone.
The timing coincides with the implementation of the Interim Measures for the Administration of Artificial Intelligence Anthropomorphic Interaction Services, effective July 15. Issued in April, the rules cover AI services that “simulate human personality traits, thinking patterns and communication styles to provide sustained emotional interaction”.
The rules exclude customer service bots, knowledge Q&A, workplace assistants, education and scientific research tools, as long as they do not involve sustained emotional interaction.
The measures cited risks including extremist ideas, privacy leaks, harm to physical and mental health, and dependence or addiction.