How can I use AI in my daily work?

‘Ghost AI’ refers to situations where employees use artificial intelligence applications in their daily work without official guidelines, creating security vulnerabilities.

AI is omnipresent and often invisible. It is used in smartphones (correction aids, facial recognition), social media (recommendation algorithms), online shopping, GPS navigation and even banking.

Ghost AI threatens security

In everyday working life, many employees already use AI – usually without thinking that this could have consequences or even legal implications. This is known as ‘ghost AI’ and can lead to serious security breaches.

A recent study by KPMG and the University of Melbourne, mentioned by Netzwoche, also looks at the dark side of AI. For example, 77% of working Swiss people regularly use AI in a professional context, which is above the global average of 58%. However, more than half of users do not comply with company guidelines, for example by uploading sensitive data to publicly accessible AI tools.

AI systems store data and continue to use it. If confidential information is entered into unsecured tools, it can be transmitted and reused. This violates data protection laws and can result in reputational damage and legal consequences.

What does the law say about AI?

In the EU, AI regulations came into force in August 2024, and many Swiss companies are already referring to them. The law aims to introduce AI applications in a safe and ethically responsible manner in practice, as well as to protect the fundamental rights of citizens.

The regulations govern AI systems using a risk-based approach: prohibited applications such as ethically questionable ‘social scoring’ (collecting assessments and behaviour of individuals) are banned, high-risk systems (e.g. justice or healthcare) are subject to strict requirements, while limited and minimal risks are only subject to transparency requirements or no requirements at all. Companies must carefully document high-risk AI and identify generative AI. High fines are planned for violations of this act.

AI in Switzerland

At the end of March 2025, Switzerland signed the Council of Europe Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law. Our country thus wishes to ratify the Convention and incorporate it into Swiss law. The aim is to promote innovation while minimising risks. A consultation draft is expected by the end of 2026.

How to use AI correctly from a legal perspective

In Switzerland, there are not yet any specific legally binding regulations on how AI tools can or must be used in everyday working life. However, employees and companies must also comply with the applicable data protection law (DPA) when using AI. The DPA applies directly to AI-assisted data processing.

What is "ghost" AI?

Ghost AI refers to artificial intelligence applications that are introduced and used in a company without the knowledge or explicit consent of the IT department.

The DPA regulates the protection of personal data and its processing. It guarantees the protection of individuals' privacy and sets requirements for companies and authorities regarding data processing.

In practice, this means:

  1. Employees must not enter sensitive or personal data into AI tools.
  2. As companies must ensure the health and safety of their employees, this also includes fundamental protection against excessive surveillance and monitoring by AI tools.
  3. In addition, employers are required to inform their employees if they use AI in exceptional cases within the permitted scope to analyse or monitor work, and to obtain the consent of employees where necessary.
  4. Employees can ask their employer at any time what data about them is being processed and for what purpose. Only necessary data about employees may be collected and processed, and only for the specified purposes. The data collected must be protected by the employer through technical and organisational measures.

Investing in safe AI

It is in the best interests of companies to establish operational guidelines and train their employees on the potential dangers of using AI. For companies in the EU, this has been mandatory since 2 February 2025. Another option is for companies to invest in secure AI models.

Discuss this topic with your colleagues and superiors if you notice a lack of awareness in your company. At Angestellte Schweiz, you will find a series of interesting courses (in German) to learn more about the safe use of AI in your everyday working life.

Author

Manuela Donati

Manuela Donati

Find out more