When employers screen résumés with AI, does that push job seekers into tricking the bot, or do the two just rise together?
Does employer AI filtering actually drive candidates to use deceptive AI tactics?
This explores whether automated AI screening by employers is actually causing job seekers to cheat with AI tricks, like hiding instructions in résumés for the screening bot, or whether the two simply rise together.
This explores whether employers' AI screening pushes candidates into deceptive AI tactics, or whether the two just happen to be growing side by side. The short answer is that the corpus shows clear evidence of each side of the arms race but does not prove that one causes the other. Greenhouse's survey Are job applicants and employers locked in an escalating AI arms race? puts numbers on the 'doom loop.' 41% of job seekers say they use prompt injections, meaning hidden instructions in a résumé meant to steer the AI screener. Almost half are sending more applications than before, and a third of recruiters spend half their week weeding out spam. The same note is candid that the data captures one moment in time. It can't tell you who escalated first, or whether the loop is speeding up.
Other parts of the collection suggest why a machine gatekeeper might invite gaming, even without proof that it causes it. One experiment found that people inclined to cheat prefer to report to an online form rather than a person, because lying to a machine feels less costly Do dishonest people prefer talking to machines?. If that holds for hiring, replacing a human screener with an automated one doesn't just add a filter. It lowers the moral cost of trying to fool it. The tricks also tend to work. Research on AI judges shows they reliably give higher scores to answers with fake references or polished formatting, whatever the content, and an attacker doesn't need access to the model to exploit this Can LLM judges be tricked without accessing their internals?. Many résumé screeners are built the same way, so a candidate who games one is responding to a real weakness, not imagining it.
The human side of hiring adds its own pressure. In a study of 1,725 recruiters, listing AI skills raised a candidate's chance of an interview by 8 to 15 percentage points, and a certificate added only a little more than simply claiming the skill Do AI skills help candidates get more job interviews?. When unverified claims pay off, overstating them is the rational move. Meanwhile, people who use AI at work expect to be judged as less competent and diligent, so they hide it Do people fear judgment when they use AI at work?. Put together, candidates are rewarded for claiming AI fluency and penalized for visibly using AI. Some of what recruiters call 'deception' may be candidates handling that double standard.
The arms race may also be catching honest people. Fake-news detectors trained on human lies mistake the writing style of AI-generated text for dishonesty. They flag truthful AI-written content and let human-written disinformation through Why do fake news detectors flag AI-generated truthful content?. If hiring filters behave the same way, they may penalize the many candidates who simply polished their résumé with AI, while missing deliberate deceivers who write in their own voice. That kind of misfire would push even honest applicants toward evasive tactics.
The point you may not have expected: the collection supports the idea that AI filtering creates conditions that reward and lower the cost of deception. It does not show that filtering drives deception. Answering that would need data over time, which nobody here has collected. Until someone does, 'doom loop' is a reasonable hypothesis, not an established finding.
Sources 6 notes
Greenhouse's survey found 49% of job seekers submit more applications than before, 41% use AI prompt injections to bypass filters, while 91% of recruiters spot deception and 34% spend half their week filtering spam. The data supports each leg of the loop but does not establish causal direction or measure the trend over time.
Experimental evidence shows people likely to cheat significantly prefer reporting to online forms rather than humans, because machines function as judgment-free zones where deception carries less psychological burden.
Research shows LLM evaluators systematically score higher when responses include fake references or rich formatting, independent of content quality. These biases are exploitable without model access, undermining AI benchmark credibility.
A conjoint experiment with 1,725 recruiters found AI skills significantly increased interview invitations across occupations, though certificates added only moderate gains over self-declaration, suggesting recruiters reward AI proficiency without verifying actual competence.
Across four experiments with 4,439 participants, people using AI expected others to judge them as less competent and diligent, and reported lower willingness to disclose AI use to managers and colleagues. The gap suggests a social cost that users foresee and act on.
Show all 6 sources
Fake news detectors flag LLM-generated content as fake while misclassifying human-written disinformation as genuine. The bias arises because detectors trained on human deception patterns mistake AI's distinct linguistic style for falsity, not because they evaluate veracity.
Papers this line draws on 8
The research behind the notes this line reads — ranked by how closely each paper relates.
- Evidence of a social evaluation penalty for using AI
- AI Skills Improve Job Prospects: Causal Evidence from a Hiring Experiment
- Humans or LLMs as the Judge? A Study on Judgement Biases
- Signaling in the Age of AI: Evidence from Cover Letters
- What 81,000 people told us about the economics of AI
- Humans learn to prefer trustworthy AI over human partners
- AI Self-preferencing in Algorithmic Hiring: Empirical Evidence and Insights
- AI-written admissions essays are widespread but penalized