"Switch language"

The EU AI Act at Telekom

3 min read

Artificial intelligence is a key driver of innovation for Telekom. We use AI purposefully to improve customer experiences, solve business challenges, increase productivity, and develop innovative products. We are committed to the safe, transparent, and ethical use of artificial intelligence (AI). We also implement the EU AI Act responsibly. 

The EU AI Act entered into force on August 1, 2024. It is the first comprehensive regulation governing AI in the European single market. Its aim is to promote innovation in a targeted way while minimizing potential risks to consumers and businesses. To support this, the AI Act establishes clear regulatory frameworks and binding standards for developing and using artificial intelligence. It categorizes AI systems according to their risk level. The requirements take effect in stages based on the level of risk: 

  • February 2025: AI practices that pose an unacceptable risk and endanger people's safety and rights (for example, manipulative systems or social scoring systems) are prohibited. 
  • August 2025: Documentation and information obligations for general-purpose AI models (General Purpose AI) apply. 
  • August 2026: Obligations for high-risk AI applications (for example, in healthcare or human resources management) must meet specific requirements. 
  • August 2027: Requirements for high-risk AI applications that are already subject to a third-party conformity assessment procedure apply.

Implementing the EU AI Act

At Telekom, a cross-segment team of experts is responsible for implementing the AI Act. A key milestone in our compliance strategy was the timely, EU-wide review of all AI applications for prohibited systems. This review was successful: There is no indication that prohibited AI is being used within Telekom or in our products.

All technologies, products, or platforms that Telekom uses or sells must undergo a process that reviews security and data privacy aspects. Since 2020, this so-called Privacy-and-Security-Assessment has included a specific digital ethics assessment for AI technologies, helping ensure effective controls for responsible AI use. The process includes clear rules for classifying, analyzing, and assessing our AI systems. 

In addition, we specifically develop employees' AI skills and raise their awareness of the responsible use of AI. Since 2018, we have offered a comprehensive and innovative range of training on the technology's potential, how it works, and its risks. This includes, for example, a gamification application on „Digitale Ethik in Aktion“ and eLearning on the EU AI Act. Prompt-A-Thons, where employees use generative AI to solve real business problems, and AI communities also help build greater awareness in working with AI. 

We are convinced: Responsible, legally compliant technology development requires clear requirements - for ourselves and our business partners. We uphold this responsibility through AI governance at Telekom. 

Background
As part of the EU AI Pact, participating companies voluntarily commit to at least three core measures:

1. Developing an AI governance strategy: To ensure responsible use of AI across the organization and work toward future compliance with the requirements of the EU AI Act.

2. Mapping high-risk AI systems: Systematically identifying AI systems classified as high-risk under the EU AI Act (EU AI Regulation).

3. Continuously promoting AI literacy and awareness among all employees to ensure ethical standards and responsible AI development.