For decades, the rigorous, multi-stage selection funnel was the “gold standard” of Talent Acquisition – a series of filters designed to isolate the true “signals” of candidate quality from the noise. We relied on the resume to screen for experience, standardized assessments to measure cognitive ability and personality, and structured interviews to verify both technical competence and job and organizational fit.

But those signals are fading in the age of generative AI (GenAI).

As GenAI becomes a standard desktop tool, candidates are no longer just using it to polish a sentence; they are using it to “co-pilot” their way through every stage of your funnel. Our research strongly suggests that the incidence of AI use is rising sharply, and it is fundamentally changing the way assessment works.

The adoption of GenAI tools has been explosive – ChatGPT’s weekly active users quadrupled from roughly 200 million to 800 million between the data collection points of our two studies. That surge in general use is spilling over into hiring. Our data shows that in late 2024, fewer than 3% of applicants reported using GenAI for assessments; by late 2025, that number jumped to nearly 19%. This is an urgent reality for every talent acquisition professional.

The “Total Funnel” Erosion

1. Resumes: The Loss of “Effort” Signals
   Traditionally, a well-structured resume signaled conscientiousness and detail-orientation. Our research found that these compositional qualities still drive hiring decisions and by a surprising margin. In a study tracking 183 job applicants through a competitive hiring process, we found that a one-point improvement in the writing quality of both resumes and cover letters was associated with 7% more interviews (relative to applications submitted) and a nearly 10-day faster time-to-hire. Crucially, these effects remained strong after controlling for applicants’ actual work experience, achievements, and GPA. In other words, writing quality predicted hiring success independently of the substantive qualifications it’s supposed to reflect. However, these data were collected in 2024 and GenAI has improved so much that it can now manufacture these signals instantly.A perfectly structured resume no longer guarantees a highly organized human; it only guarantees a user who knows how to prompt an LLM. What once signaled effort has become easy, making it harder for recruiters to distinguish between intrinsic candidate quality and AI-enabled presentation. We have heard countless anecdotes from hiring managers who are seeing larger and larger discrepancies between how candidates are presenting themselves on their resumes versus how they are performing in interviews.‍
2. The Quantitative Collapse: From GPT-4 to o1

Read the full article here: