Intelligence

在陪伴类App、AI虚拟恋人或具备情感交互能力的智能陪伴机器人爆火的今天，AI开始模拟人类特征、记忆用户偏好并与用户形成关系羁绊。这种情感陪伴式AI提升了技术温度的同时，也带来了极具挑战的问题，如关系畸形依赖、用户极端行为等心理健康问题。

2026年6月18日，国家互联网信息办公室、工业和信息化部、公安部联合公布《网络数据安全风险评估办法》（以下简称“《办法》”），并将于2026年8月20日起施行。

Chinese e-commerce platform Pinduoduo has stepped up efforts to regulate artificial intelligence-related content and services on its platform, as authorities move to strengthen oversight of the rapidly developing AI sector.

Whilst compliance audits for personal information protection are taking into shape, network data security risk assessment is now on the horizon. Is this an unbearable burden for enterprises, or can it be a convenient way to meet compliance requirements?

在上篇中，我们主要围绕AI智能体的技术特征和风险类型，分析了其在目标控制、权限调用、数据处理、工具使用、供应链管理及责任追溯等方面可能引发的安全与合规问题，并对境内外监管机构、行业组织发布的相关规范、指引及实践动向进行了梳理。

In recent years, overseas generative AI tools such as Copilot, ChatGPT, Claude, and Gemini have gradually entered the daily operations of enterprises. From drafting emails and organizing data to internal collaboration and client communication, these tools have significantly enhanced efficiency.