Employers are increasingly using recruitment software to match candidates, grade applications and screen CVs, while jobseekers are also using AI to prepare applications. Campaigners say this creates an “AI war” in which employers manage high application volumes and candidates try to make their submissions more visible to automated systems.

People Like Us is campaigning for greater transparency and accountability through its Reject the Rejections initiative. It cites a Censuswide survey of 2,000 UK workers who had applied for at least one job in the previous 24 months, in which 75% said they were unaware of their right to challenge an automated rejection. The organisation is calling for audits of recruitment systems, publication of results and independent assessment.

The same research found that 70% of ethnically diverse jobseekers had altered or concealed information such as their country of origin, religion or name because they feared rejection by recruitment software, compared with 57% of white candidates. Jobseekers interviewed by the BBC described rapid automated rejections and said they would value human review and feedback, although one said she worried that requesting a review could harm her application.

Use of AI among recruiters appears to be lower than many candidates believe. Around 90 of 425 recruiters surveyed by CV Library said they used AI during CV screening, and the company said candidates’ expectations were influencing them to write applications for bots rather than people. Recruiters also reported that AI-assisted applications often looked similar, potentially encouraging more automated filtering of large volumes of submissions.

Evidence about bias is not conclusive. Stanford researchers analysed four million applications processed by one AI hiring tool over the four years to December 2022 and found that some applicants were rejected across multiple roles more often than expected, while Black and Asian applicants were recommended at lower rates. The researchers’ findings were limited to one tool, and factors including education and immigration status may also have influenced the results. Workday is separately facing US legal claims alleging discrimination by its tools; the company denies the claims and says its products support rather than replace human judgment.

Recruitment systems vary widely, including tools that grade text-based chat interviews rather than screen CVs, making their effects difficult to assess uniformly. The UK Information Commissioner’s Office said employers should disclose when automated decision-making is used, explain how it works and identify and reduce bias; the government said existing equality and data protection laws apply and that further protections may be introduced if needed.