Job recruiters are meeting an uncanny new kind of candidate: the AI avatar. These digital humans mirror an applicant’s voice and likeness while being something else entirely, and they are turning up in first-round interviews.
The phenomenon is still new enough that many hiring managers are caught off guard. But the numbers suggest it is already widespread, and the trend is accelerating. From AI-generated application materials to deepfake video interviews, technology is reshaping the hiring process in ways that are forcing recruiters, employers, and candidates to rethink what an interview actually measures.
How one recruiter spotted it
Lana Kersanava, a Sydney-based recruiter, told Semafor she has encountered AI avatars over the past six months. “At the beginning, it’s like you feel like something is off,” she said, adding: “It looks very realistic.”
The tell was not the rendering, which held up. It was the answers, delivered with a polish that never faltered across a run of questions. Human candidates, even well-prepared ones, hesitate, stumble, or show signs of thinking. The avatars did none of that. After roughly five minutes, Kersanava ended the interview and disqualified the candidate. She has now done this three times.
Her experience offers a useful reminder: the current generation of AI avatars may not be as undetectable as the hype suggests. But the fact that she encountered three such cases in half a year suggests the practice is becoming common enough to demand attention.
The detail worth pausing on
All three avatar candidates had applied for English-speaking roles as non-native speakers, leaning on the technology to cover a gap in language proficiency. That is a different proposition from someone faking a work history. It also lands in a hiring market with a documented bias problem. Research has found recruiters up to 19% less likely to follow up with job seekers from immigrant and ethnic minority backgrounds than with equally qualified candidates.
None of which makes an undisclosed avatar acceptable. It does complicate the framing of these candidates as straightforward fraudsters. The candidates may see the avatar as a tool to overcome a barrier that has nothing to do with their actual ability to perform the job. After all, language proficiency in a professional setting is not always a reliable proxy for skill, and the hiring process has long been criticised for rewarding polish over potential.
The broader context matters too. The same technology that allows a non-native speaker to simulate fluency is also being used by candidates to automate technical interviews, fabricate experience, and game automated screening systems. The line between self-improvement and deception is shifting quickly.
The system they are responding to
Applicant tracking systems already use AI to filter CVs before a human reads them, and recruiters increasingly deploy AI avatars to conduct first-round interviews themselves. Candidates are now answering machines with machines. The resulting AI-on-AI war is rewriting the hiring playbook, pushing employers towards live technical walk-throughs, scenario challenges, and roleplay simulations that are harder to automate through.
The volumes explain the arms race. Nearly two-thirds of candidates now use AI in applications and 87% of US hiring managers use it in hiring, alongside a 239% surge in AI-generated applications. These are not marginal numbers. They show a fundamental shift in how both sides approach the hiring process.
The use of AI by candidates has exploded since the release of large language models. Tools that can write cover letters, answer interview questions in real time, or generate realistic video responses are now widely available. Some are free or very cheap, and they require little technical skill. As a result, candidates are increasingly willing to use them, whether to save time or to present a version of themselves they believe will pass the screening.
How common this is
A 2026 Greenhouse report found 91% of US hiring managers have encountered or suspected AI-generated answers during online interviews. Across 19,368 live interviews analysed between July 2025 and January 2026, 38.5% of candidates were flagged for AI-assisted behaviour. That is a strikingly high proportion, and it suggests that the problem is no longer occasional.
The trajectory points one way. Gartner projects that by 2028 one in four job candidate profiles globally will be fake, spanning AI-generated profiles, deepfake video interviews, and fabricated work histories. Barriers to entry are minimal. Palo Alto Networks found it takes as little as 70 minutes for someone with no image manipulation experience to build a fake candidate capable of passing a video interview.
These figures are likely to keep rising as the tools improve. Early deepfakes were easy to spot due to glitches, unnatural blinking, or distorted audio. Newer systems are far more convincing, especially in short video clips where there is limited time to scrutinise facial movements or voice patterns.
Employers are retreating to the room
Seventy-two per cent of recruiting leaders now run at least one in-person stage specifically to counter AI-assisted fraud, with Google, McKinsey, and Cisco reintroducing mandatory face-to-face rounds. That solves detection at the cost of everything remote hiring was meant to deliver. Candidates who cannot travel, including many of the international applicants this technology most tempts, lose access first.
The move back to in-person interviews is understandable. A face-to-face conversation is still difficult to fake convincingly, especially under sustained questioning. But it reverses years of progress in reducing hiring friction and widening the talent pool. For many employers, the trade-off is acceptable: authenticity and trust are worth more than convenience.
Yet the retreat to the room is not an option for every company. Small businesses, startups, and organisations with limited budgets may not be able to fly candidates across borders. And for global hiring, in-person interviews are often impractical. The result could be a two-track system: one for candidates who can appear in person, and another for everyone else.
The screening was never that good
There is an awkward premise underneath the panic, which is that the automated filtering being defended works. The case against AI CV screening has been made repeatedly and forcefully, largely on the grounds that it optimises for keyword matching rather than capability.
Candidates optimising against that filter with better tools is a predictable outcome rather than a moral collapse. If the machine screens the machine, the interview stops measuring anything. That is an argument for changing the process rather than policing the candidates.
The debate over AI avatars in interviews is therefore not simply about fraud or ethics. It is about the purpose of the first-round interview and whether the current process is worth saving. If a five-minute conversation can be convincingly faked by an avatar, the interview is not doing what recruiters think it does. It is measuring performance, not predictive ability.
Some employers are already adapting by focusing on skills-based assessments, portfolio reviews, and live problem-solving exercises. These methods are harder to automate because they require spontaneous, context-specific responses. They also tend to be more equitable, because they reduce the weight of charisma and accent.
Kersanava’s five-minute detection rate also suggests the current generation of avatars is not fooling attentive humans. What it is doing is making the first round worthless for everyone. A candidate who relies on an avatar may be making a rational choice in a game designed by machines. The real responsibility lies with employers to design screening systems that actually measure the qualities they care about.
As AI becomes embedded in every stage of work, the boundary between human effort and machine assistance will continue to blur. The hiring process will have to adapt, not by banning technology, but by deciding what it genuinely wants to see. The rise of the AI avatar is not just a new kind of candidate; it is a signal that the old way of hiring no longer fits the world of work being built around us.