Is browser preference a personality flaw? AI job interview evaluation raises questions
A business consultant is raising alarms about AI-conducted job interviews after he says a tech company’s evaluation of him drew some concerning conclusions, including criticizing his "habitual" use of Google's Chrome internet browser.As some companies outsource job interviews to artificial intelligence, rejected candidates can be left wondering what went wrong.After not hearing back about a job he applied for in Madrid with marketing company Anteriad, Daniel Alvarez, who is based in Spain, decided to find out exactly how the AI judged him.He obtained a copy of the AI-generated evaluation from Anteriad under the European Union’s General Data Protection Regulation. The company had used a third-party firm called ChattyHiring to conduct the screening interview. Alvarez, who is not Canadian but lived in Toronto for much of last year, shared the full evaluation and transcript with CBC News. He said he was not impressed by what he found, and doesn't feel companies should use AI interviews in the hiring process. “It's not a human-to-human interaction when you have, for example, language repair.... I can say something, and depending on your face, I can immediately rephrase it," he said. "That’s gone in this kind of interaction."Browser preference or personality flaw?Anteriad’s Dec. 22 email to Alvarez provided a link to a WhatsApp screening interview and told him not to use AI in his responses. “We value authenticity and want to hear your personal perspective,” the email read.The job was for a project involving Google. Alvarez says a bot texted him questions on WhatsApp, and he was instructed to record voice notes in response.Daniel Alvarez used EU data protection laws to obtain an AI evaluation of his recent job interview. (Submitted by Daniel Alvarez)At one point, the AI interviewer asked him which internet browser he uses daily, and why. He responded that he uses Chrome and said he does so “mostly out of habit.” The AI interviewer responded positively, saying, “Chrome is super popular for sure, easy to sync across devices!”But Alvarez later learned that the AI appeared to note his response as a strike against him, stating in its evaluation, “Habitual use of Chrome without exploring other browsers may indicate a lack of adaptability.”Jason Millar, the Canada Research Chair in the ethical engineering of robotics and AI at the University of Ottawa, called the question “absurd.”He sees the response as a “minor hallucination,” saying AI has a tendency to make things up to answer a prompt or fulfil an assigned duty. “I'm concerned about the kind of unfettered proliferation of these systems for reasons exactly like this,” Millar said. Experts who spoke with CBC News agreed that apart from that specific question, much of the AI evaluation seemed benign.Company defends AI evaluationA 2025 study by job board Indeed found that 87 per cent of hiring leaders in Canada are using AI in the hiring process, for tasks that could include writing job descriptions, summarizing resumes and interviewing candidates. ChattyHiring CEO Carlos Guerrero says criticism of Alvarez's evaluation is misguided.In an email to CBC News, he said no scoring or weighting was assigned to the Chrome criticism, despite it being noted as a "con" in a "pros and cons" list. Guerrero said even a score of zero will generate a list of "pros" from the AI tool, while a perfect 10 will still generate a list of "cons" — though he noted "some of them might sound strange." "Their purpose is simply to provide additional perspectives for recruiters, not to influence the evaluation criteria defined by the employer," he said.Guerrero said the companies hiring — in this case, Anteriad — design the interviews and evaluation criteria, while his firm conducts the interviews and relays the results, without making the final decision.Anteriad’s legal department told Alvarez in an email, which Alvarez shared with CBC News, that a human reviews all automated results before making hiring decisions. The company did not respond to CBC News’ request for comment.A Google spokesperson said in an email to CBC News that all of its hiring decisions are made by human recruiters.AI interviews reduce HR workloadAI interview companies market themselves as a way for businesses to cut labour costs, hire better employees and reduce workload for their HR teams. Hilke Schellmann, an investigative journalist and author of The Algorithm: How AI Decides Who Gets Hired, Monitored, Promoted and Fired, And Why We Need To Fight Back, says the proliferation of AI interviews comes as companies are complaining about a “deluge” of resumes.She says this has been an issue since the advent of two-click applications on platforms like LinkedIn. Now that AI agents can also apply to jobs for candidates, she says some HR departments are faced with the “arduous” task of going through seemingly impossible numbers of resumes.WATCH | Young Canadians weigh in on AI:CBC Kids News asked young people in Toronto and Barrie, Ontario, and in Edmonton, Alberta, how they feel about artificial intelligence — and the answers were mixed. From excitement about careers and innovation to concerns about jobs, the environment, and relying too much on technology, this streeter explores how youth are thinking about AI and the future.Proponents argue that in addition to saving time, the technology makes better decisions by eliminating human bias.But Schellmann says that's not always how things play out in practice, because the systems are set up to do statistical analysis without an ethical or moral compass of their own. Companies also tend to be secretive about their internal systems and are mostly left to audit themselves if problems do arise, she said. “It can bring a new bias, and it can replicate the bias that's already in the training data, which we see again and again,” Schellmann said. A collective action lawsuit in the U.S. alleges HR software company Workday’s AI system discriminated against older candidates, an allegation Workday has denied. Back in 2018, Amazon scrapped an automated job candidate ranking tool after finding it effectively favoured men over women. Schellmann says the interviewing process itself is flawed, even when only humans are involved, often rewarding confidence over competence. And while AI did not create those flaws, she says it can in some cases exacerbate them."I think we've automated a pretty poor process to begin with."Data concernsApart from the evaluation itself, Alvarez was upset by the mere involvement of a third-party company, something he said was unexpected and that left him with questions about who has his data now.Guerrero says in this case, ChattyHiring is the data processor, while Anteriad is the data owner. He says ChattyHiring's interviews are "fully private" and his company only collects the candidate's first name and contact information. Though many Canadian companies use AI interviews, experts told CBC News it would be complicated, if not impossible, under Canadian laws for an interviewee to obtain the data Alvarez did from a private company.Jason Millar, Canada Research Chair in the ethical engineering of robotics and AI at the University of Ottawa, has concerns about the proliferation of AI job interviews. (Submitted by Jason Millar)Millar also has concerns about data security with AI interviews, with new companies popping up all the time and no standards in place in Canada.While European companies are subject to stricter privacy rules and regulations, Millar would like to see stronger protections for Canadian candidates' rights, and says people should have the right to opt out of AI interviews. “Are the productivity gains here worth it?" he said. "Or is this dehumanizing?”
