AI recruitment tools are discriminating against refugees and migrants, according to new research.

Deakin University researcher Bis Hakimi said that AI powered job application platforms were biased against foreign sounding names.

“Our research show AI algorithms are filtering resumes and CVs and reinforcing unfair outcomes,” Mr Hakimi told the recent Refugee Communities Association of Australia (RCAA) conference in Adelaide.

He said that one point of discrimination was in the fact that many refugees did not have linear career records because they may have spent time in a refugee camp.

“The technology is not neutral, and the design of the technology auto amplifies and shapes who gets hired and who gets excluded,” said Mr Hakimi, a research assistant and PhD candidate.

“The question is how we build technology that is more inclusive and more adaptable to people’s lived experience,” he said.

Mr Hakimi said he had experimented in applying for jobs using his name and his supervisor’s Anglo-Saxon name as the sample control.

“The results were skewed way in favour of someone with an Anglo name. And we found that David and Michael were the names most favoured by AI tools,” he said.     

“AI systems are only as unbiased as the data they’re trained on. To ensure fairness, organisations should regularly audit their AI tools and incorporate diverse datasets during development,” Mr Hakimi said.

“While AI enhances efficiency, it should never replace the empathy and intuition of human recruiters. A balanced approach ensures that technology supports, rather than overshadows, the human element of recruitment,” he said.

A study last year also found refugees in Australia find it harder to get decent jobs that other migrants.

The main reason was employer discrimination rather than a lack of skills or language proficiency, the research by the University of NSW said.

Lead researcher and economics lecturer Dr Liwen Guo said that humanitarian migrants in Australia remain significantly less likely to be employed than non-humanitarian migrants, even after five years of resettlement.

This is largely due to employer discrimination rather than any shortfall in skills or language, she said.

The study, published in the British Journal of Industrial Relations, showed differences in education and language may explain some of the employment disadvantage faced by refugees in Australia, but the main factor is discrimination by employers because to how they screen candidates when they lack reliable information about overseas qualifications or experience.

“Although relying on visible traits like race, age, or appearance may conflict with Australia’s anti-discrimination laws, this kind of behaviour is usually subtle and informal. Employers do not say openly that they are rejecting someone because of their background; instead, when they are unsure how to judge overseas qualifications or experience, they fall back on ‘gut feeling’ and visible cues,” Dr Guo said.