The new AI regulations are here. What action do businesses urgently need to take?

New AI Rules – urgent action by companies required: from now on, all companies that use or develop AI systems such as DeepL, Microsoft Copilot, ChatGPT, corporate GPTs, and Adobe Firefly must provide AI training to their staff and other relevant individuals. In addition, as of 2 February 2025, certain AI practices are prohibited:

1. AI literacy (Art. 4 AI Act):
- Companies must ensure that employees and other individuals working with AI systems on their behalf have received adequate training by 2 February 2025.
- Non-compliance may lead to fines and compensation claims. These legal consequences are not explicitly stated in the AI Act but, precisely because of this, they pose considerable risks. Companies can reduce their exposure proactively by implementing
the correct measures.
2. Prohibited AI practices (Art. 5 AI Act):
- The AI Act prohibits certain AI practices deemed by the EU to pose unacceptable risks. This raises critical questions: can employee sentiment analysis based on email evaluations be conducted? What facial recognition or biometric identification and
classification systems will remain legal? What qualifies as ‘exploitative and manipulative AI practices’? Where does social scoring start?
- Non-compliance can lead to draconian penalties of up to €35 million or 7% of global annual turnover.

Three actions you should take immediately:
1. Take stock of all AI systems used in your company and evaluate them for any prohibited practices.
2. Implement training programmes and initiatives to build AI skills.
3. Establish a robust AI governance system. Effective governance not only reduces legal risks but also empowers your team and drives innovation.

We understand the challenges that companies will be facing. If you have any questions, don’t hesitate to contact our partner Dr. Frederic Geber.
#AIAct #AIGovernance #Compliance #Literacy #AICompetence
AI Park