Focused on People:
Telefónica Deutschland subscribes to six ethical principles regarding artificial intelligence
"The Telefónica Group has declared worldwide that it will only use artificial intelligence in accordance with explicitly formulated ethical principles", said Telefónica Deutschland CEO Markus Haas at the Telefónica BASECAMP in Berlin on 13 February 2019. Telefónica Deutschland has interpreted these principles for the local market. They are intended to give the company direction when it is working with data and artificial intelligence, thereby supporting the positive impact of technology on society. Artificial intelligence (AI) is a key element in the digitalisation process in Germany. This was recently underscored by the AI strategy adopted by the German government. People and organisations rely on AI to process large volumes of data on the basis of complex algorithms in order to enhance decision-making and realise automation solutions. There is a wide variety of examples that illustrate how AI leads to improvements in the economic and societal situation. These include improved medical diagnoses, driver assistance systems, optimised supply chains and disaster relief support. Everyday applications, like smartphone speech-to-text functionality and intelligent thermostats, also employ AI to some extent.
Telecommunications firms bear great responsibility
As a telecommunications enterprise, we are involved in this development and are, in fact, at the core of it. After all, we not only enable digital applications, we also use algorithms and artificial intelligence ourselves – for instance, for the optimisation of our mobile network – in the form of the digital assistant Aura or for data analyses for traffic planning. Accordingly, we bear great responsibility towards the economy and society. Telefónica has thus declared that the company will exclusively utilise artificial intelligence in accordance with explicitly formulated ethical principles. These are standards for our company, but also must be observed by our business partners and suppliers as well. These principles are additionally intended to stimulate public discourse. They are an integral part of our Responsible Business Plan. "The ethical principles we have adopted on the use of artificial intelligence place the focus on people. At Telefónica Deutschland we will apply these ethical principles in artificial intelligence projects”, insists Valentina Daiber, Chief Officer for Legal and Corporate Affairs at Telefónica Deutschland.
Six principles of ethics for the use of artificial intelligence
I. “Fair and accessible to all – digitalisation is non-exclusionary”.
It is important to us that technological applications based on artificial intelligence yield ethically proper results. We aim to ensure that they in no way affect people in a discriminatory way in relation to race, ethnicity, religion, gender, sexual orientation, disability or other personal circumstances. Furthermore, all members of society should have access to digital technologies. This also means that everyone should be empowered to use such technologies and assess the relevant developments as part of digitisation.
II. “AI should serve society. We will only be able to overcome the prevailing scepticism within society if artificial intelligence delivers concrete personal advantages”.
AI should serve society and yield concrete benefits for people. To this end, AI systems must remain under human control and be managed under a values-based approach. In utilising AI in products and services, we will uphold human rights and remain in alignment with the United Nations Sustainable Development Goals. In those areas where the human brain reaches its limits or can be freed of routine tasks, AI delivers real added value. It is essential that the individual derives personal benefit. In order to mitigate reservations about AI, it is important to make its usage transparent through understandable communication.
III. “Not everything that is possible is morally legitimate – boundaries must be set for artificial intelligence”.
Society and the business sector have to cooperate in setting clear boundaries for the proliferation of autonomous systems and AI – even internationally. We promote societal dialogue on this complex of issues. Additionally, before the use of AI at Telefónica, we review whether doing so is ethically justifiable, discussing the potential long-term impact.
IV. "The use of data needs to be transparent and these data have to be protected".
We believe every individual should have the control to decide regarding the usage of his/her data in accordance with our Data Protection Policy. The prerequisite for this is transparent information. However, AI transparency means more than data protection. Users must be informed whenever they are interacting with an AI system. We ensure transparency on whether a customer is interacting with our chatbot or a person. It is important to know, after all, whether an outcome results from man or machine so that users can assess what it is based on. In the case of an AI-based decision, we strive to ensure that the logic behind a system decision is comprehensible to people.
V. “Data stemming from partnerships with third parties must be able to pass checks”.
AI-based products and services may be developed in cooperation with business partners or other third parties. When this is the case, we reserve the right to check the lawfulness and correctness of the data and verify the underlying logic.
VI. “Human beings must remain the ultimate instance of ethical judgement”.
We control machines; they do not control us. Ethical decisions should always be reserved for human beings. Our moral compass rests upon contextual knowledge that is not always explainable or programmable. In this respect, we are far ahead of machines. Thus in future as well, it should only be people who are allowed to decide on ethical and moral questions. And that is how we do it at Telefónica Deutschland.