Artificial intelligence guidelines

Artificial Intelligence is a hot topic. For many it is the Next Big Thing in Business. Others, such as business pioneer Elon Musk are warning about the dangers. Deutsche Telekom is one of the first companies that has responded to this challenge and defined guidelines for dealing with artificial intelligence (AI). These guidelines are self-binding guidelines with an emphasis on our business and  business models. They define how we want to deal with AI, and how we will develop our AI-based products and services.

Our customers of the Telekom Data Intelligence Hub are the first to benefit. Particularly, for a service offering that is explicitly focused on benefits from Data Analytics the guidelines are of utmost importance and mandatory for the development of our product offerings.

Artificial intelligence must support our actions. AI should support people and expand their abilities instead of restricting them.

1. We take responsibility.
   Clear definition of who is responsible for which AI system.
2. We handle artificial intelligence carefully. 
   Our AI systems and their use follow the laws and regulations that apply to people.
3. We focus on our customers. 
   AI should simplify and enrich the lives of our customers.
4. We stand for transparency.
   Transparency when a customer communicates with an AI system and transparency about the use of customer data.
5. We offer security.
   Our customer data is protected against unauthorized external access (data security and data protection).
6. We lay the foundation.
   Thorough analysis and evaluations as a basis for the further development and continuous improvement of our AI systems.
7. We keep the overview.
   Constant readiness to intervene in our AI systems in order to avoid or reduce damage.
8. We live the cooperation model.
   Although we see the human being in the first place, advantages can be derived from an interaction of human being and machine.
9. We share and explain.
   Dissemination of knowledge about AI and transfer of relevant competences.

Of course, we do not claim that our guidelines are a silver bullet, but we do want to develop our policy further through continuing discussions that will help us achieve a broad consensus.

We look forward to broad feedback and a good discussion. Please find our detailed guidelines HERE.

by

Dr. Karsten Schweichhart

18.10.2018

artificial intelligence compliance