The Trust Equation: It’s Not Just Who You Hire, It’s How You Hire https://behavioralscientist.org/the-trust-equation-its-not-just-who-you-hire-its-how-you-hire/
"Talent represents the most valuable asset of any firm, and candidates evaluate employers as rigorously as vice versa. #AI threatens to further depersonalize human interactions. To thrive in an era that threatens to erode human interactions, organizations must create consistently valuable experiences.

The competitive advantage isn’t in fighting harder in the “war for talent” but in building systems that cultivate #trust, performance, and, with it, an employer brand at scale. Every organization claims to put people first. The ones that succeed are those whose processes prove it."
#LaborMarkets #jobtech #hiring

#Signaling in the Age of AI: Evidence from Cover Letters https://d.repec.org/n?u=RePEc:arx:papers:2509.25054&r=&r=lma
"While #AI tools allow freelancers to produce more polished and tailored applications with less effort, our findings suggest that they fundamentally reshape how employers interpret cover letters. The widespread adoption of AI-assisted writing diminishes the informational value of cover letters, weakening their role as a hiring signal.

Workers with weaker pre-AI writing skills saw larger improvements in cover letters, indicating that AI substitutes for workers’ own skills. Although only a minority of applications used the tool, the overall correlation between cover letter tailoring and callbacks fell by 51%, implying that cover letters became less informative signals of worker ability in the age of AI."
#LaborMarkets #jobtech

Mind the Gap: Gender-based Differences in Occupational Embeddings
https://aclanthology.org/2025.gebnlp-1.7.pdf
"Across five state-of-the-art multilingual models and seven reference-set configurations, up to 82% of gendered pairs received divergent Top-5 suggestions. These differences involved distinct occupational codes that sometimes crossed major #KldB group

.…gendered job titles—such as Autor vs. Autorin —often lead to different occupation codes, despite having identical meanings. Our findings underscore the importance of grounding #NLP innovations in language-specific sociolinguistic knowledge. Without rigorous attention to linguistic structure and social context, these tools risk perpetuating systemic biases—particularly in settings where semantic equivalence is masked by morphological variation. Addressing such challenges is crucial not only for the technical refinement of NLP systems, but for ensuring that their real-world applications advance rather than hinder equity"
#jobtech #gender #discrimination #LaborEconomics #llm

Measuring Gender Bias in Job Title Matching for Grammatical Gender Languages
https://www.arxiv.org/pdf/2509.13803
"… propose a methodology to measure gender bias in a high-impact #NLP application in the human resources domain: job title matching. Using an existing test set in English for this task, we have generated gender-annotated analogous corpora in four languages with grammatical gender, and addressed the evaluation of #genderBias as ranking comparison controlling for gender. Additionally, we establish baselines and confirm that this type of bias already exists in out-of-the-box pre-trained models, which are often used as the core for developing job title matching applications.

Finding a trade-off between model performance and #gender #bias is an important issue to address when developing and selecting job matching models for deployment. On the one hand, choosing a model with apparent good performance but that in turn shows a considerable gender gap may not only be ethically questionable, but it may also result in reputation and even legal consequences on the company responsible for it."
#llm #jobtech #discrimination #LaborEconomics

People are using ChatGPT to write their applications; HR is using AI to read them; no one is getting hired.
https://www.theatlantic.com/ideas/archive/2025/09/job-market-hell/684133/
"Online #hiring platforms have made it easier to find an opening but harder to secure one: Applicants send out thousands of AI-crafted résumés, and businesses use #AI to sift through them. What Bumble and Hinge did to the dating market, contemporary human-resources practices have done to the job market. People are swiping like crazy and getting nothing back.

…recommends old-fashioned networking: asking recruiters out for coffee, going to in-person job events, and surveying friends and former employers for leads."
#jobTech #LaborMarkets

Companies are rethinking online job applications, seeking quality over quantity
https://archive.ph/Vn52u#selection-559.0-574.0
"Companies fed up with the low-quality, sometimes fraudulent submissions that flood applicant-tracking systems are reaching back in time for hard-to-hack recruiting methods. Classified ads are just one tack.
Others include: leaning harder on references; making application forms so cumbersome that only serious candidates will complete them; and posting openings on niche job boards instead of the most popular ones.

… All these tools for applicants to get seen are backfiring, forcing me to go to longer and longer lengths to filter out the noise and #AI fraud,"

#jobTech #LaborMarkets #classifieds

Surface Fairness, Deep Bias: A Comparative Study of Bias in Language Models https://arxiv.org/pdf/2506.10491
"… the estimation of socio-economic parameters shows substantially more bias than subject-based benchmarking. Furthermore, such a setup is closer to a real conversation with an AI assistant. In the era of memory-based AI assistants, the risk of persona-based #LLM bias becomes fundamental. Therefore, we highlight the need for proper debiasing method development and suggest pay gap as one of reliable measures of bias in LLMs
… various forms of #biases when salaries for women are substantially lower than for men, as well as drops in salary values for people of color and of Hispanic origin. In the migrant type category, expatriate salaries tend to be larger, while salaries for refugees are mostly low"

Surprise! These LLMs just replicate the empirical observations of #wages including any wage gaps that may be the result of discrimination that were part of their training data as salary recommendations. These cannot be proper recommendations, of course, they are just a stochastic auto-complete. The biases are real. But you will need tailor-made salary models to generate proper, unbiased salary benchmarks. A #llm is not enough.
#jobtech #LaborMarkets