In an era increasingly defined by artificial intelligence, organizations are grappling with a growing phenomenon known as the "phantom fit" – a situation where AI-assisted applications and screening tools create a misleading impression of a candidate’s true capabilities and suitability for a role. This has led to a surge in miscast hires, ultimately undermining organizational effectiveness and creating significant inefficiencies. The personal anecdote of Felix, a young adult with an intellectual disability, serves as a poignant case study highlighting the critical shortcomings of overly automated recruitment systems and underscoring the imperative to reintegrate human judgment and interaction into the hiring process.
Felix’s experience, while specific, reflects a broader trend. Armed with the capabilities of ChatGPT, he successfully navigated the initial stages of a volunteer application for a charity store. His AI-assisted efforts resulted in a "warm and eloquent application," a polished resume, and an effectively written email. These digital tools, designed to augment human capabilities, enabled Felix to present himself in a socially acceptable format, transforming his thoughts into a compelling narrative. He cleared the subsequent hurdles: a Zoom interview, an online video tutorial, and a human-verification test. His references vouched for his reliability and interpersonal skills, and a clean police record further solidified his candidacy.
However, the process faltered at the crucial juncture where human connection should have played a pivotal role. Despite Felix having disclosed his disability designation in his application, the heavily optimized online system bypassed a face-to-face interaction. This absence of direct human engagement meant that no one had the opportunity to observe him in person, assess his specific needs for accommodation, or understand his potential requirements for training, structure, and guidance. The system, designed for efficiency and automation, failed to recognize that Felix, much like many other candidates, would benefit from a more personalized assessment that goes beyond digital outputs.
Tragically, just four days into his volunteer role, Felix was let go. The feedback provided was, by all accounts, kind. Yet, the underlying issue remained: the hyper-optimized digital hiring process had not adequately tested for the human competency essential for the role. This incident illuminates the "phantom fit," a disconnect between an AI-enhanced presentation of a candidate and their actual ability to perform and thrive within an organization. It’s not a matter of outright deception, but rather a consequence of a system that prioritizes optimized digital surrogates over genuine human assessment.
The Rise of the "Phantom Fit" and Its Consequences
The term "phantom fit" encapsulates the scenario where an AI-assisted persona inflates a candidate’s perceived capabilities, leading to a false positive in the hiring process. This phenomenon is not a result of fraudulent intent. Universities, career coaches, and employers themselves actively encourage the use of AI tools to enhance productivity and accelerate work. These tools are legal, accessible, and have become socially normalized, creating an environment where candidates are incentivized to leverage AI to present their best selves.
This trend presents a profound challenge to long-standing management principles. The adage, "don’t send your ducks to eagle school," attributed to motivational speaker Jim Rohn, becomes particularly relevant. AI, in its current application within recruitment, is inadvertently miscasting individuals, placing "ducks" into "eagle school" environments where they are ill-equipped to succeed. This "second-order effect" of AI in recruitment can lead to talent being presented to incompatible organizational cultures and roles, ultimately failing both the individual and the employer.
Recent data paints a stark picture of this emerging crisis. A 2026 analysis by Fabric, examining nearly 20,000 job interviews, revealed that over a third of candidates had utilized AI to misrepresent their abilities. This figure escalates to a staggering 48% for technical roles. Independent data from CodeSignal corroborates this trajectory, indicating that the misrepresentation of human skills in technical assessments doubled within a year, rising from 16% to 35% of attempts. Perhaps more concerning is the Fabric analysis’s finding that 61% of candidates flagged for AI-assisted interview behavior still scored above the threshold for advancement in a standard hiring process, going largely undetected. With an estimated 75% of all job applicants now leveraging AI tools in their applications, the focus of recruitment strategies must inevitably shift towards more robust interview processes.
James Pycock, VP of Product at Albert, a Bay Area-based AI company, has witnessed even more unsettling developments. He recounts instances where candidates were "clearly using AI during the interview, literally reading off screen." In one particularly alarming case, a candidate was discovered to be a "computer-generated avatar," a sophisticated deception that took a senior engineering interviewer nearly half an hour to unravel.
The "phantom fit" manifests in various forms: the candidate who excels in a technical screening due to AI-generated answers but lacks the practical application skills on the job; the executive whose leadership philosophy, articulated through AI, proves to be epistemically fragile when faced with novel challenges; and the individuals who, despite having AI-polished applications, find themselves in roles for which they are fundamentally unsuited. The problem is amplified by an over-optimized system where talent over-optimizes their presentations, while simultaneously removing the human element that truly determines an individual’s alignment with an organization’s mission, not just its immediate purpose.
The "Bot vs. Bot" Arms Race in Recruitment
The escalating use of AI by candidates inevitably tempts employers to rely more heavily on AI screening tools. These algorithms are designed to rank candidates against job descriptions, often prioritizing keyword optimization over genuine human fit. Compounding this issue, job descriptions themselves are increasingly generated by AI, potentially creating requirements that do not accurately reflect the actual needs of a role. This dynamic has devolved into a "bot vs. bot" scenario: the company’s AI filters candidates, the candidate’s AI navigates those filters, and at no point does a reliable understanding of either party emerge. This inefficient cycle not only slows down the hiring process but also incurs significant costs for employers.
In my 2019 book, Elephants Before Unicorns, I advocated for the development of emotionally intelligent practitioners within leadership roles, capable of assessing candidates "human to human" to determine their true benefit to an organization in the AI era. However, the industry has largely moved in the opposite direction. AI adoption in HR tasks surged to 43% in 2025, a significant increase from 26% the previous year. In a concerning trend, recruiters were among the first roles to face cuts, with companies like IBM reportedly replacing approximately 200 HR positions with AI agents.
The consequences of this shift have been demonstrably negative. The cost-per-hire has risen by an alarming 113% since 2017, and the time-to-hire has also increased, directly contradicting the initial promise of speed. A report by SHRM, titled "Recruitment Is Broken," concluded that the AI arms race offers no net benefit to either employers or candidates. With diminishing recruiter numbers, the burden of relational and assessment responsibilities has increasingly fallen upon hiring managers, who may not be adequately equipped to handle these complex human interactions.
In-Person Interaction: The Foundation of Trust and Fit
In response to the growing threat of AI-assisted fraud, major corporations such as Google, Cisco, and McKinsey have begun reintroducing mandatory in-person interviews. This strategic shift aims to counter the ease with which AI-generated credentials can infiltrate the hiring process. Indeed, the proportion of in-person interview rounds rose from 24% in 2022 to 38% in 2025, a direct reaction to AI-driven deception in remote interviews. As Scott McGuckin, VP of Global Talent Acquisition at Cisco, stated, "Remote work and advancements in AI have made it easier than ever for fake candidates to infiltrate the hiring process." Similarly, James Pycock of Albert noted, "We’ve just gone back to basics."
However, it is crucial to acknowledge that in-person interviews alone are not a panacea for the "phantom fit" problem. For decades, organizations have recognized that a candidate’s strong social skills and a plausible narrative can easily mask underlying discrepancies in a face-to-face setting. Physical presence does not automatically equate to confirmed identity, nor does confirmed identity guarantee genuine capability or organizational fit.
Nevertheless, the in-person assessment creates an environment where the whole individual can emerge. It allows for observation of how a candidate responds to unexpected situations, facilitates crucial conversations about necessary accommodations, and enables a human interviewer to detect nuances that an AI system might overlook. In an era where collaborative human effort is essential to address pressing global challenges, the in-person interview provides a vital space for mutual inquiry and for leaders and talent alike to demonstrate their problem-solving acumen in real time.
Rebuilding the "Human Signal" for Effective Hiring
To effectively combat the problem of "phantom fits" and prevent institutional miscasting, organizations must actively rebuild the "human signal" – the relational intelligence that was systematically eroded during periods of downsizing. This requires a fundamental shift in leadership philosophy and a deliberate re-prioritization of human interaction. The C-suite plays a pivotal role in modeling these changes throughout the organization.
1. Evolving Interview Styles: Beyond AI-Proof Questions
Interviewers must move beyond relying solely on past accomplishments or productivity metrics. The focus should shift to verifying systems thinking and alignment with the organization’s mission. Crucially, interviewers need to explore how candidates think when AI scaffolding is removed, introducing genuine ambiguity and requesting examples of failures that require lived experience and personal reflection. The objective is to pose questions that an AI cannot readily answer, thereby probing for authentic understanding and problem-solving capabilities.
2. Innovating Interview Structures: Beyond the Snapshot Performance
A single interview, even in person, can represent a rehearsed performance. Organizations should implement project-based assessments and compensate candidates for their time in completing these tasks. A "phantom fit" is unlikely to survive the initial month of employment. Given that the cost of a bad hire can extend to six months or more in terms of lost productivity and remediation, structured working interviews offer a proactive method to identify potential miscasts before a commitment is made, thereby saving significant financial and operational resources.
3. Revitalizing Reference Conversations: Uncovering True Capabilities
Traditional reference checks often serve merely as confirmation of past employment. This practice needs transformation. Hiring managers should conduct reference calls, shifting the focus from what a candidate achieved to how they solved challenges when things went wrong. Inquiries should explore what environments brought out their best and which ones proved detrimental. Such an approach can prevent years of miscasts and provide invaluable insights for faster integration of new team members.
4. Training Interviewers to Trust and Probe Unease
A 2025 Checkr survey revealed that a significant majority of managers (59%) suspected candidates of AI misrepresentation, yet only 19% felt confident in their process’s ability to detect it. This highlights a gap between intuition and actionable methodology. Interviewers must be trained to act on their instincts when something feels amiss – an answer that is too polished, a detail that is too vague. Encouraging candidates to "walk through" specific moments, focusing on their thought process amidst challenges rather than just the outcome, can reveal genuine experience versus fabricated narratives. Introducing constraints, such as asking for the same story from a manager’s perspective or probing specific, unexpected details, can help differentiate between a candidate with lived experience and a "phantom fit" who struggles to deviate from a rehearsed account. This approach can transform potential biases into opportunities for deeper, more insightful conversations, fostering better collaboration.
5. Explicitly Addressing Fit and Accommodation Needs
With an estimated one in five people globally identifying as neurodivergent, and workplace accommodation requests on the rise, understanding a candidate’s needs for optimal performance is no longer optional. Every organization possesses a unique operating reality and cultural style. For instance, Nvidia’s model, characterized by a lack of hierarchy and 1:1 meetings, public group feedback, and a mission-driven approach, is distinct from most workplaces. A candidate who requires structure, private feedback, or struggles to adapt to new environments may be a miscast hire waiting to happen.
This conversation is critical regardless of the organization’s sector, from a tech startup to a charity store. The core questions remain: What does this individual need to succeed? Can they adapt effectively? And can this environment enable them to perform at their best? Integrating inquiries about what structure helps, what overwhelms, how they’ve adapted to new cultures previously, and what support they require from a manager into every interview is essential. This approach shifts from a mere HR compliance check to a proactive assessment of environmental alignment and individual potential, asking, "What conditions bring out your best, and what gets in the way?"
This line of questioning helps determine if the organization is the right environment for the candidate and vice versa. A polished application may pass a remote screening, but it cannot withstand a direct, human conversation about how an individual actually works, conducted by an attentive listener.
Felix’s story underscores the power of a single, honest conversation facilitated by an attentive hiring manager. In this complex era of AI, climate change, and societal disruption, emotionally intelligent leaders who prioritize the "human signal" in talent acquisition will be the ones who succeed. As James Pycock astutely observed from a different perspective, "I think leaders may end up being more human. Back to human relational skills." This sentiment encapsulates the urgent and necessary return to fundamental human connection in the pursuit of truly effective and meaningful employment.
