An article by Thomas Kremer, Board member for Data Privacy, Legal Affairs and Compliance.
The German government is developing a strategy for the utilization of Artificial Intelligence (AI). Chancellery minister Helge Braun recently stated that today, the problem was not the restriction but rather the enablement of AI. We at Deutsche Telekom share this view. Therefore, we have given ourselves guidelines for the ethical handling as well as the application and development of AI.
Likely, many people still consider AI to be science fiction. But it’s not. At Deutsche Telekom, we are already using AI. For example, a service bot answers simple customer requests. Our CONNECT app uses intelligent machine learning to know the quality of the available networks. Thus, the app can decide for each location whether Wi-Fi or the mobile network offers the better connection. AI can even assist with maintenance work for bridges. A digital early warning system for bridges, tunnels or buildings composed of various sensors detects warning signals early before any visible indications occur.
You see, AI exists in various kinds. But three things can always be applied – the so-called ABC of AI:
A: Algorithms – only through programming machines become intelligent.
B: Big Data – data is the feed for the algorithm. The machine can learn only from data.
C: Computing power –processing and transporting billions of data sets needs enormous computing power and fast networks. This is where we as Deutsche Telekom come into play. The new network standard 5G will even increase the opportunities in this field.
In the future, AI will become even smarter through better algorithms, it will become even more precise through more available data and it will become even faster through best networks. A real power technology of the future. But such a powerful technology has to be tamed in the right balance. It has to be given a conscience and awareness for ethical behavior. Not everything which is technically possible, should also be allowed. Now that this technology is spreading, it’s the right time to set the right course. We must shape the modus operandi we would like to apply to AI now.
This rationale is becoming more and more commonplace. The call for ethical policies for AI is eligible and becomes louder. In politics, science and economics. In Germany, Europe and the United States.
- The German coalition agreement plans a commission for data ethics.
- German Federal Minister of Justice and Consumer Protection Katarina Barley wants to collaborate with IT companies to create ethical guidelines, a Corporate Digital Responsibility, for the digital world.
- Yesterday, Google’s CEO Sundar Pichai published six guidelines for the ethical utilization of AI.
- The EU commission wants to establish an “on demand” platform and an observatory for AI to simplify “the access to the latest algorithms”. Moreover, they are working on a Charta of Ethics (planned for 2019).
- Leading scientists suggested the founding of a European Lab for AI in an open letter in April.
So, there are a lot of activities around this topic. At Deutsche Telekom, we act now. Already since the beginning of last year we are discussing this topic with our internal AI experts and specialists from the Silicon Valley. This year, in the beginning of May, we published nine guidelines for the application and utilization of AI. We are dedicated to these guidelines since then. It starts with a clear commitment for the responsibility of humans. This means: There is one person in charge for each KI system in our company. This person can always interfere in AI systems to prevent damage. And we commit to transparency: Customers always know whether they are communicating with a human or AI. We disclose which customer data we use for AI systems and protect them against unauthorized, external access. These are just some examples. You can read all of our nine guidelines here.
Our guidelines do not want to be a “philosophers' stone” – even though we developed them with all our AI experts within Telekom and other external experts as well. We want to discuss our guidelines. With you, the public, with politics and with other companies. That is why I also discussed the guidelines with our Data Privacy Board. The members affirmed us in how important it is to introduce such guidelines as a company. And that the guidelines are good ones. They were positively surprised that we are among the first companies to actively address this topic. The discussion was intense and encouraging. A good sign that we are on the right track. Let’s develop and shape this power technology of the future together. I look forward to more feedback and good discussions.