AI recruitment machines map out workers’ career paths

AI is playing an increasing role in determining career paths Keystone

Artificial intelligence (AI) is not only changing the nature of work, it’s changing who does the work, with tools that influence hirings, firings and promotions. This has raised fears that self-learning machines will dictate career paths.

+ Get the most important news from Switzerland in your inbox

Companies, recruiters and job seekers are already experimenting with the technology to gain an edge. Algorithms can write job descriptions and filter applications to find the best candidate or help build training programmes by identifying skills gaps among current workers.

AI can also help job seekers write CVs and covering letters, and some apps prompt candidates to give the right answers during remote interviews and tests.

But there is a growing recognition of a tipping point where the use of AI becomes more of a hindrance than a help. At the centre of any work relationship are people, who need to connect and use their own intuition when making decisions.

Even Silicon Valley AI giant Anthropic has asked job candidates to limit the use of technology when applying for positions, according to the Financial Times.

“A battle of machine versus machine is not helping solve the challenges of labour and skill shortages,” Liana Melchenko, board member of the ISACA Foundation, set up in 2020 to promote diversity in the technology workforce, told SWI swissinfo.ch.

“Employers are starting to recognise this and are paying more attention to creative thinking, complex problem solving, the ability to learn and emotional intelligence. A job applicant’s most valuable qualities are what makes them human.”

Fundamental change of work

AI has also led to uncomfortable questions about the confidentiality of personal data and the ability of machines to treat staff with dignity and empathy.

+ Young Swiss seek jobs less prone to AI disruption

There have been examples of badly programmed AI systems replicating discrimination, such as a recruitment tool that was mothballed by Amazon in 2018 after showing bias against women.

And the use of AI to judge peoples’ emotions from video footage has stirred up misgivings about the intrusion of machines into sensitive personal areas.

Policy makers around the globe, including Switzerland, are debating how to control the impact of AI on the recruitment and monitoring of employees.

“AI is fundamentally changing the world of work. Stricter regulation is essential to protect workers’ rights and to guarantee them a right to consultation when AI systems are introduced,” said Daniel Hügli, a board member and head of ICT at the Swiss trade union Syndicom.

In February, Syndicom demanded that Switzerland adopts European Union regulations on the use of AI in the workplace. This includes data protection measures, full disclosure of when and how AI is used and a ban on systems that analyse peoples’ emotions.

Companies adopting AI

Switzerland is meandering down its own slow, measured path towards AI regulation. The Alpine state has signed up to the Council of Europe AI Convention that safeguards human rights. But it will take at least two years before Swiss laws are adapted accordingly.

A Swiss communications ministry report, published in February, points out that any Swiss company or agency working in the European Union realm would automatically be subject to the tight regulations of its AI Act, which came into force in 2024.

+ Why China’s approach to AI intrigues Switzerland

AI tools are just starting to make their presence felt in the Swiss workplace. A SWI swissinfo.ch survey of Swiss companies, including supermarket chain Migros, pharmaceutical firm Novartis, state-owned telecommunications company Swisscom and food producer Nestlé, found firms adopting a gradual and cautious approach to the technology.

For example, Migros, Switzerland’s largest retailer, has developed a chatbot on WhatsApp for school-leavers to ask questions about apprenticeship posts. This tool allows people to find out more about apprenticeship offers but it plays no role in determining which candidates are selected, Migros says.

Advantages of AI

“Our recruitment philosophy remains to prioritise the human element of the hiring process,” a spokesperson for Nestlé said. “Where AI solutions are identified, we follow a rigorous process to ensure that they are fit for purpose.”

But while companies are generally playing it safe, they are also alive to the efficiency enhancements that AI can offer when utilised responsibly.

Swiss recruitment firm Adecco, for example, has recently doubled down on its use of AI-powered recruitment software from United States technology company Bullhorn.

AI recruitment software can swiftly spot patterns in large volumes of data that prove elusive to the human eye. The systems can then join the dots to suggest enhancements in a two-way ‘conversation’ with their human handlers.

Companies and workers should also seize on the advantages offered by well-trained algorithms, according to Ghislaine Couvreur, who was CEO of the now-defunct Swiss AI platform Vima Link until May 2022.

Machines should serve people

Vima Link specialised in identifying perceived personality traits and soft skills, primarily for recruitment. The company failed to get off the ground due to governance issues and went into liquidation in 2023.

+ Swiss workers less protected from AI than EU counterparts

Couvreur acknowledges that fully complying with EU restrictions, including the new AI Act, which took effect in August 2024, would have potentially posed an additional challenge to the platform.

But she remains convinced that the technology could bring positive benefits if it is carefully built and contains enough checks and balances, such as the ethics board established by Vima Link to mitigate risks and provide appropriate guidance.

“This can definitively help employees and managers better understand how they are performing and how they are perceived by others, while removing bias as much as possible.”

“AI systems only work when they serve people and enhance their experiences,” she told SWI swissinfo.ch. “This can help employees better understand how they are performing and how they are perceived by others.”

But convincing workers that such systems are objective is another challenge for AI developers to tackle.

Edited by Gabe Bullard/sb