AI辅助面试在秋招季大规模应用，企业借助算法进行批量初筛以提升招聘效率1。这类面试工具主要用于第一轮筛选，采用录播半结构化模式，通过语音转写、语义分析、面部特征等手段进行打分1。然而，该做法引发了从业者和求职者的广泛质疑。
中国人事科学研究院研究员庞诗指出，"AI面试工具目前还不是成熟的人才测评工具，最后还是要通过HR或业务负责人进行判断"1。某高校就业平台负责人孙鹏在试用半年后选择停止使用，他表示"AI面试不太精准，有时会出现误伤"1。HR总监黄涛则指出AI无法准确评估"积极性、进取性"等软特质，并警示"一旦被求职者摸清打分逻辑，就可以针对性做面试准备"1。
业内专家普遍认为，AI面试不应作为唯一淘汰依据。庞诗强调，"目前市场上还没有统一权威的AI面试工具和行业标准，建立相关规范已经十分紧迫"，并认为需要明确适用范围、数据采集边界和算法透明度要求1。求职者也普遍反映缺乏真人互动、评分规则不透明、隐私顾虑等问题1。
AI-powered interview screening has seen widespread deployment during China's autumn recruitment season, with companies leveraging algorithmic systems to conduct mass initial screenings and improve hiring efficiency 1. These AI interview tools typically operate in a recorded, semi-structured format, analyzing candidates through speech-to-text conversion, semantic analysis, and facial feature recognition to generate scores 1.
However, significant concerns have emerged regarding the reliability and transparency of this approach. Researchers and industry practitioners have raised critical questions about the technology's maturity. Pang Shi, a researcher at the China Institute of Personnel Science, stated that "AI interview tools are not yet mature talent assessment instruments, and final decisions still need to be made by HR personnel or business managers" 1. A university employment platform director, Sun Peng, discontinued use of AI interviews after a six-month trial, citing that "the AI screening was not sufficiently accurate and sometimes produced incorrect eliminations" 1.
Beyond accuracy concerns, the limitations extend to evaluating soft skills. Huang Tao, an HR director, pointed out that AI cannot effectively assess qualities like "initiative and ambition," and warned that "once job seekers figure out the scoring logic, they can prepare their interviews strategically" 1. Industry experts also highlight the absence of standardized solutions. Pang Shi emphasized that "the market currently lacks a unified, authoritative AI interview tool and industry standard, making the establishment of relevant regulations extremely urgent," calling for clear guidelines on applicable scope, data collection boundaries, and algorithm transparency requirements 1.
评论
还没有评论，欢迎留下第一条。