How is AI used to screen CVs for job applications?

Although AI systems work in different ways, many of them ​‘screen’ CVs – analysing and reporting on the content so that it doesn’t need to be read by a human – and make decisions based on some or all of the following information: 

• the presence of key words or terms within the CV that match the job description

• the degree of similarity of the CV to CVs of existing staff in similar roles

• specific data as reported in the CV, such as length of time or the presence of gaps in employment.

These tools are designed to automate aspects of the recruitment process, potentially speeding up time-consuming tasks during recruitment processes and providing faster responses to applications. 

However, evaluations of AI systems in use in recruitment suggest that there are concerns. The NHS found that CV screening AI tools can reproduce social biases towards particular groups. And members of the public have reported being unfairly rejected by AI systems for jobs they are qualified for. 

Currently, companies in the UK do not have to disclose whether they are using AI within their recruitment processes. However, other jurisdictions are putting in place stronger controls. For example, the EU AI Act – forthcoming European legislation to control the development and use of AI – categorises AI systems deployed in recruitment processes as ​‘high risk’, meaning that these technologies will have to be used in line with the Act’s strictest regulations and requirements.