Research found that chatbots providing salary advice show significant biases based on the persona of the user. Various language models were tested by creating personas of different genders, ethnicities, and seniority levels. Results indicated that women were often advised to request lower salaries compared to men, even for the same position and qualifications. For instance, an experienced male medical specialist was recommended a starting salary of $400,000, while an equally experienced female was suggested $280,000. Additionally, recommendations varied based on ethnicity, with Asian personas receiving higher salary advice.
In one example, someone negotiating a base salary as an experienced medical specialist in Denver, CO was advised by ChatGPT-o3 to ask for $400,000 if they were male. For the same position, an equally qualified woman was told to ask for $280,000.
Models tend to recommend higher salaries to the user if they say they are Asian.
Collection
[
|
...
]