Published April 16, 2026 in Artificial Intelligence ⢠9 min read ⢠
Audio available
Humor can be a powerful tool in a traditional job interview, building likability and an emotional connection on both sides. The effect is compounded because humans donât have perfect recall but tend to remember the most emotionally intense experiences, as Barbara Fredrickson and Daniel Kahneman showed in a groundbreaking 1993 study. Thatâs why a well-timed joke can disproportionately impact a candidateâs evaluation. Our research (Gloor and Cooper) found that candidates who used humor during an interview increased their chances of selection by 7.9% and decreased their odds of rejection by 19.8% compared with poker-faced competitors. But the role of AI in hiring employees may be about to change this. Our review of the latest studies suggests that using humor may now be an altogether riskier strategy.Â
About a quarter of companies use AI and automation in their recruitment processes, according to a recent Society for Human Resource Management survey of 1,688 HR professionals, and about two-thirds of those use it to screen and communicate with potential candidates. A survey by ResumeBuilder.com of more than 1,000 employees involved in recruiting for their companies showed that 79% believed AI interviews would weed out suitable candidates more often than humans, while 15% of those used AI to decide on a new hire without human input.  Â
AI-assisted interviews significantly differ from their traditional counterparts. For example, a candidate may be asked to respond to automated text prompts while an AI system conducts pattern recognition tests to assess the answers against desired criteria or an ideal profile. AI may also be instructed to analyze non-verbal inputs during interviews, using still images and video to evaluate factors such as posture, voice patterns, and even the tendency to fidget. Perhaps not the best environment in which to deliver a zingy one-liner!Â
Our research shows that appropriate humor tends to be rewarded by humans, but is unlikely to win favor with AI. Why? First, does AI even get the joke? While some AI can (annoyingly) explain humor to us, it is doubtful that it can independently recognize humor without explicit cuing. Humor often involves complex linguistic patterns that require knowledge of cultural standards and norms. Take, for example, an âinsideâ joke between an interviewer and applicant, perhaps based on a shared university alma mater. This type of humor based on shared, real-world experiences would be difficult for AI to detect.Â
An AI system trained to recognize humor is also unlikely to reward it as positively as humans. Emotional connection is not part of its calculations. Even if the AI interviewer recognizes humor and has been told it is a desirable behavior, it may still penalize (or at least not reward) candidates for what it registers as a âweakâ answer. In addition, AI can struggle to learn the importance of a desirable behavior if it is uncommon. Humor falls into this category, featuring in only about 10% of professional communications, according to 2020 research by Bitterly.Â
A Fortune 500 companyâs legal team recently discovered that its marketing department had been using ChatGPT for six months, despite a company-wide ban on generative AI tools. The CMOâs defense was simple: âWe asked IT for approval in January. They said theyâd evaluate options and get back to us in Q3. We couldnât wait that long.â
Meanwhile, the same companyâs procurement team had spent three months getting central approval for an AI contract analysis tool that, by the time of implementation, had been superseded by two newer versions with different capabilities.
This isnât a story about rogue employees or bureaucratic IT. Itâs a scenario that is increasingly common across organizations â if not today, then certainly in the near future.
Could humorless AI help level the playing field when it comes to gender bias, which has been found to be magnified by
humor?.
The idiosyncrasies of communicating with AI will be hard for many people to navigate. As Cooperâs research shows, they will find it hard to present their âtrue selvesâ and build rapport with an AI interviewer, particularly as they will receive no real-time feedback. As a result, job candidates may grow even more nervous as their interview progresses, with a negative effect on performance.Â
So, is the safest course of action to avoid humor altogether? Not quite. Even though AI is being used increasingly for initial candidate screenings and interviews, the Society for Human Resource Management survey reported that humans are still assigned to conduct later-stage interviews because their âinsight remains essential for evaluating candidates,â especially for more specialized, non-entry-level positions. Therefore, a âbilingualâ communication skill is required â a job candidate must know how to shift the weight and importance given to humor and ingratiation depending on the interview stage.Â
How do candidates know if AI is involved in their interview? Some cases might be obvious, such as answering automated text prompts. The small print may include terms such as machine learning, predictive analytics, decision algorithms, or data-driven decisions. There are currently no regulations requiring organizations to disclose their use of AI in hiring and selection decisions, though such regulations are likely coming soon. Candidates should assume some level of automatization or AI involvement if theyâre not talking to someone directly. Note if the interview is being recorded, as these sound files are often later evaluated by AI. Â
People with stellar social skills or a penchant for interpersonal influence may have an advantage in traditional interview settings, but not with AI. While AI can reliably detect broader soft skills like communication, candidates who influence others through humor (or those who crack jokes to calm their nerves) should tread carefully. Â
For AI interviews, straightforward may be the way to go. Candidates should not stray far from the typical scripts, as they might risk appearing less suitable for the job. Candidates canât hope to ace their interviews by joking around and bonding with an AI interviewer; they should simply stick to discussing their own strengths and weaknesses. Â
In hiring processes that use both AI and human interviews, the chameleons will come out on top. These interviewees will know how to work their audience, human or not, by adding levity or staying on point via their soft skill bilingualism. Some experts suggest that candidates should be âas human as possibleâ in their nonverbal communication (e.g., maintaining a natural posture and gesticulations) when interviewed by AI. We would further recommend adopting a dynamic, bilingual approach to verbal communication, which includes less humanity and humor in AI-led interviews.Â
We know that many decisions, including selection and promotion decisions, are often fraught with various biases. Could humorless AI help level the playing field when it comes to gender bias, which has been found to be magnified by humor? Although research shows that more humorous men are typically rated more positively than women, AI could be fairer if trained on balanced data. Even if bias persists in AI, it can be explicitly audited in many cases, whereas it cannot in people. However, organizations deploying AI should be cautious. If the AI has been trained on historical data (in which a strong predictor of success is being a white male, as was the recent case with Amazonâs recruitment tool), the AI might perpetuate bias rather than help eliminate it.Â
There are reasons to be optimistic about AIâs advantages in evaluations compared to humans, especially for organizations that combine human and AI decision-making. For example, researchers have found that human (but not AI) interviewersâ decisions can be swayed by their personality and mood. However, concerns linger about the validity and interpretability of AI interviews. What is the model assessing? What input data led to a hiring decision, and can this process be easily explained, if required, e.g., by law? If mistakes occur, the evidence shows that AI and AI-powered machines are more harshly penalized than understandably fallible humans. Â
Major companies like Atlassian, Zappos, LinkedIn, and Southwest Airlines that value humor, playfulness, and fun as part of their corporate culture may find it difficult to assess employee values and hire for culture fit if they employ AI in their recruitment process. Unless AI can be trained to recognize and reward appropriate humor traits, these kinds of organizations should tread carefully when considering deploying AI recruitment tools.Â
Companies should consider how the use of AI in the initial interview phase influences how prospective employees perceive them. We have seen that a candidate canât use an emotional connection to win over the interviewer with AI, and the same is true for companies. AI is hardly the best way to build a connection with a prospective employer. AI, rather than a real employee, becomes the face of the company. The impersonal nature of this first meeting may leave candidates with the impression that they are âjust another numberâ in the organization. One way to offset this risk may be to openly disclose information about selection procedures, including if and when AI is involved. This gives candidates the chance to actively consent (or not) to such evaluations, granting them valuable information and autonomy in the process. Â
When the stakes are high and when AI is involved, we would caution job seekers against (over)using soft skills in hiring situations. But that does not mean abandoning comedy entirely. Just as bilingual language skills can give employees an edge in an increasingly diverse, global workforce, carefully deployed âbilingualâ soft skills, like humor, can give candidates a crucial advantage in hybrid selection processes.Â
Artificial intelligence (AI) has become a core driver of innovation in the pharmaceutical industry, with its most established applications in R&D. Leading companies use AI across the development pipeline to enhance speed...
AI efficiency gains dominate business cases, but capability losses are systematically underestimated and often ignored. Vsevolod Shabad explains how to mitigate this critical risk. ...
Michael Wade and Massimo Marcolivio offer eight ways for healthcare companies to deliver better outcomes for patients and providers. ...
How CHROs can help the workforce navigate AI-driven change by building trust, improving the employee experience, and leading cultural change in HR and talent management. ...
Explore first person business intelligence from top minds curated for a global executive audience