Abstract: The core objective of empowering elderly care services through intelligent assistive technology lies in achieving both precision and humanization of care by reconstructing service delivery via technological mediation. At its essence, this transformation represents a dynamic balance between algorithmic power and human-centered values. The evolutionary trajectory of technological logic indicates that while the integration of Internet of Things (IoT) architectures and multimodal sensing systems can enhance service efficiency, it also gives rise to emerging legal risks—such as the substantive failure of informed consent mechanisms due to cognitive decline in the elderly, the privacy threats posed by pervasive surveillance technologies, and the difficulty in identifying clear legal responsibility among actors. In response, legal regulation must construct a dual-adjustment mechanism of “technology-adapted law” and “law-guided technology”: on the operational level, a layered informed consent framework for aging populations should be established, with simplified interfaces and strengthened algorithmic interpretation rights to safeguard decision autonomy; in the data governance dimension, a dynamic “right to be forgotten” system should be introduced, along with innovative “machine forgetting” technical standards to ensure control over the data lifecycle. In terms of legal responsibility, a multi-layered legal responsibility system should be constructed, with a “responsibility chain allocation mechanism” based on technological processes and a “gradient responsibility system for risk sharing” to allocate liability proportionate to the ability to control risks and the benefits accrued.

Key words: intelligent assistive technology, elderly care services, legal risks, legal regulation, algorithmic power