同学你好，该知识点来自沪江网校《2025年11月CATTI笔译二级【签约班】》的课程，想要更系统的学习，欢迎进入课程学习。不仅可以和更多的同学一起学习，而且还有老师、助教随时的学习指导和知识点解答哦。

同学好~

首先为同学的勤奋点赞！

2. 第二段（你的译文）
We should advocate building a system to test and assess layered risks and exercise swift governance, classified and graded management and effectively rapid response.
The R&D entities need to constantly make AI more explainable and predictable, make data truer and more accurate.
By this means we can ensure that AI is always under human's control and could be verified, supervised, traceable and reliable.
主要错误
layered risks → risk-level（固定搭配）
swiftly governance → agile governance（敏捷治理 = 官方固定译法）
effectively rapid → effective and timely
make data truer → enhance data authenticity and accuracy
human's control → human control
could be → is
3. 第三段（你的译文）
We need to gradually establish and improve legislation and regulations to guarantee the security of individual privacy and data throughout AI development and application
and counteract any behavior aiming to poach, falsify, disclose and illegally collect and utilize others' private information.
主要错误
individual privacy → personal privacy
poach（偷猎）→ misappropriate, steal
非法收集利用 → illegal collection and use（固定）
4. 第四段（你的译文）
We adhere to the principle of equity and indiscrimination so as to avoid generating any prejudice and discrimination targeted to different or special ethnic, beliefs, nationalities and gender
amid the process of data obtainment, algorithm design, technology developing and products research, developing and application.
致命扣分点
indiscrimination → non-discrimination（完全错）
targeted to → against
ethnic（形容词）→ ethnic groups
gender → genders
data obtainment → data acquisition
technology developing → technology development
products research, developing → product R&D and application

希望帮到同学~