The use of artificial intelligence (AI) is still a relatively new technical phenomenon but it is already having a fundamental impact on all areas of economic life. This also applies to human resources (HR) – where various companies are already using AI. This is part 13 of our AI blog series.
Possible uses of AI in human resources
AI is already having a certain impact in HR and can change the way HR experts organize their work. It is conceivable that AI could be used in the following areas in particular:
Recruiting and talent acquisition:
- Generative AI can create better job descriptions that clearly define all required skills.
- Automatically screening CVs against these specifications to create shortlists that are reviewed by humans;
- Conducting preliminary interviews, e.g. through short questionnaires, to gather information that can then be used by human interviewers;
- Generating relevant, personalized and unbiased questions for interviewers;
- Analysing interviews (text form, recordings or videos) to broaden context for hiring decisions;
- Sending personalized responses to applicants throughout the application process.
Personnel development and training:
- AI can create personalized training plans and adapt learning content to the individual needs of employees;
- Automated learning pathways and recommendations for further training based on employees' skills and interests.
Employee commitment and retention:
- AI can analyse employee feedback and recognize trends to identify areas for improvement;
- Personalized recommendations for career development and training.
Performance Management:
- AI-based tools can monitor employee performance and automatically generate notifications for performance reviews.
Management of employee data:
- Automated data collection and processing, e.g. when hiring or pomoting employees.
Advantages of using AI in human resources
The use of AI in human resources (HR) offers a number of advantages:
Increased efficiency:
- AI can automate repetitive tasks, such as screening CVs or managing employee data. This gives HR experts more time for strategic tasks.
Better basis for decision making:
- AI analyses large amounts of data and identifies patterns. This enables informed hiring, promotion and performance appraisal decisions.
Personalisation:
- AI can create personalized training plans and provide employees with individual training recommendations;
- Personalized career development suggestions based on individual skills and interests.
Objectivity and equal treatment
- AI makes decisions based on data and algorithms, without human bias.
- AI can reduce the risk of discrimination and promote equal opportunities.
Talent acquisition and retention:
- AI helps identify talent that fits the company's goals;
- Improved employee engagement through personalized approaches.
Time saving:
- Automation of processes such as appointments, onboarding und offboarding.
Improved employee experience:
- AI-based chatbots can help employees with questions and concerns around the clock.
Performance management:
- AI can monitor employee performance and provide early warning of issues.
Overall, the use of AI can help make HR departments more effective and data-driven, which can ultimately lead to a better working environment for everyone.
Disadvantages and risks of using AI in human resources
The use of AI in HR entails a number of risks and potential disadvantages, including the following:
Human alienation and lack of transparency:
- If AI systems are used too heavily, this can lead to alienation between employees and the HR department. Personal interactions could be reduced, which can have a negative impact on employee engagement.
- AI cannot replace human interaction. In areas such as recruiting, personal contact between applicants and HR managers is crucial.
- Many AI models are so-called "black boxes". This means that their decision-making process is difficult for humans to understand. However, transparency is important in the HR sector in order to build trust and ensure fair decisions.
Bias and discrimination:
- AI algorithms learn from historical data and can adopt unconscious biases. If this data is discriminatory, the AI can also make discriminatory decisions;
- It is important to regularly review and improve AI models to minimise such biases.
Privacy and security:
- The use of AI requires access to and the processing of large amounts of personal data. This also carries a risk of data leaks or unauthorised access. It is crucial to ensure that this data is adequately protected to avoid data protection breaches.
Lack of human intuition:
- AI can analyse data, but it does not have human intuition or emotional intelligence. In complex situations, the human factor is (still) irreplaceable.
Technical errors or system failures:
- AI systems are susceptible to technical errors or system failures. A faulty system can lead to incorrect decisions.
Resistance to change:
- Employees may show resistance to the introduction of AI, especially if they fear that their jobs will be jeopardised by automation.
Costs and Effort required for implementation:
- The implementation of AI systems requires investment in technology, training and the customisation of work processes.
It is important to take these disadvantages into account and to
carefully plan the use of AI in HR and implement transparent, data
protection-compliant processes in order to make the most of the
benefits while minimising potential risks.
Legal Framework
Although the widespread use of AI is a relatively recent phenomenon, the use of AI is already subject to legal requirements, especially in the HR sector (For further information: Part 3: AI: 11 principles – and an AI policy for employees - Vischer ). Data protection in particular plays a decisive role in the use of artificial intelligence - data protection law is applicable to AI-supported processing of personal data (on data protection in the HR context, see also here: New Data Protection Act: What do I need to know as an HR manager? - Vischer ). Here are some important aspects:
Information:
- Employees or applicants must be adequately informed before their data is processed by AI tools. Companies should be transparent and clearly communicate how AI is used in the HR department.
Transparency:
- AI models are often complex systems whose decision-making is difficult for humans to understand. Nevertheless, transparency is important in order to gain the trust of employees. Under certain circumstances, companies should disclose how AI is used in HR and which data sources are used.
Automated decisions:
- The full automation of decisions in the employment relationship is limited. Particularly in the selection or evaluation of employees, automated individual decisions that have significant adverse effects are not unproblematic and are subject to higher transparency requirements. Human judgement is essential for important automated individual decisions. There is still a debate about how to deal with this when AI only prepares a decision.
Data protection impact assessment:
- A data protection impact assessment is required for certain AI applications in the HR sector. This assessment evaluates possible risks of data processing for the data subject.
Data security:
- The use of AI requires the processing of large amounts of personal data. Companies must ensure that this data is stored securely and protected against unauthorised access. This also applies if providers are used - appropriate contracts are required for them. Data protection breaches can have considerable legal and financial consequences.
Rights of data subjects:
- Employees and applicants have the right to information about the processing of their data, the right to rectification of incorrect data and the right to erasure or anonymization. Companies must ensure that these rights are respected.
Avoidance of discrimination:
- AI models may have unconscious biases. Companies must ensure that AI systems do not make discriminatory decisions based on gender, ethnicity or other protected characteristics.
To summarise, data protection in HR is crucial to ensure fair, transparent and privacy-compliant processes.
In addition to data protection law, the EU AI Regulation (AI Act) could also have an impact on Switzerland. The AI Act is applicable if an AI system is deployed within the EU or its output is "used" in the EU. This applies, for example, if a forecast, recommendation or decision made by an AI-based system in Switzerland is "used" within the EU. Systems that are used to assess employees and applicants are normally considered "high-risk" AI systems and are therefore subject to additional requirements. Under the AI Act, for example, it is not possible to simply screen applicants. In this case, however, many Swiss companies do not use the output in the EU. Further information on the AI Act can be found here: Part 7: The EU AI Act - what it means in practice for most companies - Vischer
In Switzerland, the Federal Administration is currently evaluating various approaches for the regulation of AI through to the end of 2024. The Federal Council will then decide on these and, if necessary, issue a corresponding regulatory mandate.
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.