90%雇主拒绝AI味简历,半数却用AI面试

一幅未来感十足的低多边形风格画面:两只手在紫色渐变背景前相握,象征商业与科技行业中的协作、信任与数字互联。sankai / iStock / Getty Images Plus via Getty Images
Greenhouse 的新数据显示,对 AI 的疑虑正在如何影响招聘。
尽管如此,企业表示它们正用 AI 来筛选更多候选人。
在低迷的就业市场中,求职者可能比平时更愿意借助 AI。
招聘经理如今比以往任何时候都更怀疑求职申请中使用了 AI,这正在影响他们考虑哪些人。与此同时,他们又转向 AI 来帮助筛选更多候选人——这意味着你在网上看到的那些求职者戏耍 AI 面试官的视频,短期内不会消失。
招聘系统公司 Greenhouse 的新数据发现,“九成美国雇主(90%)认为,他们曾因为申请看起来是假的或由 AI 生成而拒绝了真实候选人。”ZDNET 已请 Greenhouse 澄清,这 90% 的雇主是将申请送入了 AI 检测工具,还是仅凭直觉判断这些申请看起来是假的。
雇主也在通过调整招聘流程来应对 AI:86% 的受访者表示已作出改变以应对 AI 的影响。其中 57% 的人增加了一轮人工面试。
报告指出:“随着求职者越来越依赖 AI 来润色和投递求职申请,雇主越来越难判断谁才真正适合这个岗位。”这也呼应了求职者与招聘团队之间持续存在的拉锯。
AI 简历筛选器会自动丢弃大多数它认为不合适的申请,因此申请人会试图针对筛选器调整自己的材料格式以免被丢弃,但结果往往还是适得其反。
报告显示,艰难的经济形势与大量涌入的 AI 招聘工具正在相互碰撞,可以说这种碰撞更有利于雇主而非申请人。Greenhouse 指出,讽刺的是,AI 工具在帮助招聘经理的同时,也让简历变得不可信,而申请人宁愿被 AI 面试,也不愿完全没有面试机会。
“越来越多的雇主使用 AI 不仅是为了筛掉候选人,也是为了通过 AI 语音技术更早地与更多候选人交流,”Greenhouse 表示,并补充说,“现在有一半雇主使用 AI 面试来接触更多候选人,其中 40% 因此找到了原本会错过的优秀候选人。”
我们不知道、或许也无法知道的是,AI 筛选器是否曾错误地影响招聘经理,让他们拒绝掉本应予以考虑的候选人。
该报告在 9 月底至 10 月中旬调查了美国、英国、德国、澳大利亚和爱尔兰的 1,347 名招聘经理和招聘人员,发现使用语音 AI 的雇主中有一半“表示早期阶段的候选人体验有所改善”。
不过,目前尚不清楚这里的“改善”具体指什么,也不清楚申请人自己是否会认同这一点。尽管如此,根据 Greenhouse 早前的一份报告,如果可以选择,美国只有 7% 的申请人会拒绝 AI 面试。此外,该报告称,“经历过实时 AI 面试的人中,70% 表示这有助于他们展示自己的技能。”
但这在多大程度上受到美国就业市场状况的影响?10月的就业报告好于预期,但对求职者来说,整体而言今年依然是艰难的一年。Greenhouse 的发现既可能反映出求职者有多迫切,也可能代表着他们对 AI 筛选的认可或接受。
“对大多数求职者来说,与人对话从来就不在选项之中,”Greenhouse 语音 AI 负责人 Ophir Samson 表示。“真正的选择是简历石沉大海,还是用 15 分钟讲述自己的故事。每位候选人都应该得到后者。”
招聘经理仍在亲自处理的部分
Greenhouse 发现,虽然雇主正在用 AI 筛选更多候选人,但多数雇主在招聘流程中最重要的决策上仍保留人工监督。
Greenhouse 表示:“在雇主愿意交给 AI 智能体的所有任务中,决定哪些候选人进入下一轮排在最后,仅占 7%。”
加州近期的一项法律旨在保护员工,免遭 AI 系统在没有人工共同签署评估的情况下就对其绩效作出判定并据此解雇。与职场中许多 AI 生产力工具一样,若采用得当,理想情况下应意味着人类员工有更多时间处理更重要的工作。
Greenhouse 表示:“在借助 AI 节省了时间的雇主中,85% 把省下的时间重新投入到候选人身上,最常见的做法是花更多时间与入围候选人交流(45%),以及更早地面试更多候选人(40%)。”
围绕 AI 面试官的主要问题仍然是,它们能否像你期望的人类面试官那样准确捕捉你的面试表现(尽管这本身就极其主观,而且无论如何也不在求职者的掌控之中)。
“如今,每100份申请中只有3份能进入面试,而增加更多人工面试也无法弥合这一差距,”Greenhouse CEO 兼联合创始人 Daniel Chait 表示。“解决办法是扩大漏斗顶端,向每一位申请者开放语音 AI 面试,让更多人更快获得一次真正的初次对话。”
考虑到 Greenhouse 自身就提供语音 AI 筛选面试工具,他的立场并不令人意外。
最终,更完整的图景或许需要将 Greenhouse 的数据与成功通过 AI 面试官和人工面试官两关后被录用的申请者人数进行对比。
译文由英文原文机器翻译,术语按行业惯例统一,可在文末展开原文核对。
查看英文原文(90% of employers reject resumes that look like AI – yet half use AI to interview)
A futuristic, low-poly handshake between two hands against a purple gradient, symbolizing collaboration, trust, and digital networking in business and tech industries. sankai / iStock / Getty Images Plus via Getty Images
New Greenhouse data shows how AI suspicion is impacting hiring.
Still, companies say they’re using AI to screen more candidates.
A tough job market may mean candidates are more willing to engage with AI than they would otherwise.
Hiring managers are more suspicious than ever of AI use in applications, and it’s impacting who they consider. At the same time, they’re turning to AI to help screen more candidates — meaning those videos you’ve seen on the internet of people pranking AI interviewers aren’t going away anytime soon.
New data from hiring systems firm Greenhouse found that “nine in ten US employers (90%) believe they’ve turned away real candidates because those applications seemed fake or AI-generated.” ZDNET has asked Greenhouse to clarify whether that 90% of employers reported putting applications through an AI detection tool, or are just intuiting that those applications seem fake.
Employers are also reacting to AI by adjusting their hiring processes: 86% of those surveyed reported changes to account for AI’s impact. 57% of them did so by adding another human interview.
“As jobseekers rely more on AI to polish and send out their job applications, employers are finding it harder to tell who’s actually right for the role,” the report stated, echoing an ongoing push-and-pull between applicants and hiring teams.
AI resume screeners automatically toss a majority of applications they deem unfit, so applicants try not to get tossed by formatting their material for the screener, which then backfires anyway.
The report shows how a tough economy and an influx of AI hiring tools are colliding, arguably to benefit employers over applicants. Greenhouse noted that, ironically, AI tools are helping hiring managers while also making resumes untrustworthy, and applicants would rather be interviewed by AI than not at all.
“A growing number of employers are using AI not just to filter candidates out, but to have conversations with more candidates, earlier, through AI-powered voice technology,” Greenhouse said, adding that “half now use AI interviews to hear from more candidates, and 40% of those have found strong candidates they would have missed.”
What we don’t, and maybe can’t, know is whether AI screeners have incorrectly influenced hiring managers to turn away candidates they actually should have considered.
The report surveyed 1,347 hiring managers and recruiters across the US, UK, Germany, Australia, and Ireland from late September to mid-October, and found that half of employers using voice AI “say the early-stage candidate experience has improved.”
It’s unclear what “improved” means there, though, and whether applicants themselves would agree with that. Still, when given the choice, only 7% of applicants in the US would turn down an AI interview, per an earlier Greenhouse report. Additionally, “70% of those who’ve had a live AI interview say it helped them show their skills,” that report said.
But how swayed is that by the state of the US job market? October’s job report was better than expected, but it’s still been a tough year overall for job-seekers. Greenhouse’s findings could be as much a reflection of how desperate applicants are as an endorsement of or comfort with AI screenings.
“For most applicants, a human conversation was never on the table,” said Greenhouse’s Head of Voice AI Ophir Samson. “The real choice is a resume lost in a pile, or 15 minutes to tell their story. Every candidate should get the second.”
What hiring managers are still handling
Greenhouse found that while employers are using AI to screen more candidates, most are maintaining human oversight over the most important decisions in the hiring process.
“Of all the tasks employers would hand to an AI agent, deciding which candidates to advance came last, at just 7%,” Greenhouse said.
A recent California law aims to protect employees from being fired by AI systems that make determinations about their performance without a human co-signing the assessment. As with many AI productivity tools in the workplace, proper adoption ideally equates to more time for human workers to handle more important tasks.
“Of employers who’ve saved time with AI, 85% put it back into candidates, most often by spending more time with shortlisted candidates (45%) and interviewing more candidates earlier (40%),” Greenhouse said.
The main question around AI interviewers remains whether they’re correctly capturing your interview the way you’d hope a human interviewer would (though that’s still deeply subjective and out of applicants’ control anyway).
“Today, only three in 100 applications ever reach an interview, and adding more human interviews can’t close that gap,” Greenhouse CEO and co-founder Daniel Chait said. “The fix is scaling the top of the funnel by opening up a voice AI interview to every applicant, letting more people get a real first conversation sooner.”
His stance is unsurprising given that Greenhouse itself offers a voice AI screening interview tool.
Ultimately, a fuller picture might compare Greenhouse’s data with the number of applicants hired after successfully passing both an AI and a human interviewer.
Please accept cookies to view the comments powered by Disqus.
Radhika Rajkumar is a senior editor at ZDNET based in New York City. She covers AI, specializing in safety, privacy and security, policy, education, and synthetic media. She also leads ZDNET's newsletter strategy. Radhika holds a Masters in Creative Publishing and Critical Journalism from The New School. See full bio
原文为英文,正文由自动抓取并保留原文。
原文发布于 ZDNet(90% of employers reject resumes that look like AI – yet half use AI to interview),标题、摘要与正文由机器翻译,术语按行业惯例统一,未经人工逐句校订。