COMPANIES are increasingly using artificial intelligence (AI) to help identify potential job candidates and match them with the right companies. “We rely on the power of data,” says Timm Lochmann, who is responsible for everything having to do with automated data at Stepstone, an online jobs platform. He and his colleagues think the traditional protocol of jobseekers sifting through advertisements and employers hiring headhunters is outdated. “We want to find the perfect matches,” Stepstone boss Sebastian Dettmers says. “We are practically a dating service in the professional arena.”
With the help of automated data processing, Stepstone tries to match applicants with individual advertisements, without having to actively search for them. If a company places a job advertisement on Stepstone, the internal AI will present a handful of suitable candidates directly to the company. And applicants are shown a custom assortment of job listings that the AI has determined are relevant to them. This involves evaluating all the data the user creates, including voluntary personal details, past searches, even ads or websites that users click on.
Simplifying the job search
“Simplifying the job search is an extremely important topic in my opinion – especially in times of a shortage of skilled workers,” says Andreas Pinkwart, minister of economic affairs in the German state of says North Rhine-Westphalia. The fact that many people are unhappy in their current jobs shows that there is a need to improve the system. Pinkwart sees AI solutions as a “possible solution to a problem that is becoming increasingly virulent”.
Large corporations in particular have high hopes for AI-powered recruiting, but a case at Amazon shows that this can also backfire. Based on an evaluation of past successful recruitments, an AI programme identified only male applicants from select universities as candidates. Women and other people who did not conform to this profile fell through the cracks.
This can put an employer in a tricky spot. “This is a major problem when using AI: How do you train it? The learning process is very important,” says Wolfgang Brickwedde, who heads the Institute for Competitive Recruiting in Heidelberg. If an applicant feels discriminated against and a company cannot justify the rejection because AI made the decision, this is a serious problem. Lochmann from Stepstone is aware of the problem: They’re trying to anticipate such risks and consciously counteract them in the development of their artificial intelligence systems.