From production to product, from sales to the legal department: Artificial Intelligence (AI) is used in the automotive industry in a wide variety of areas. In this article, Renata Jungo Brüngger, Member of the Board of Management of Daimler AG, Integrity and Legal Affairs, explains how Daimler uses AI responsibly.
The impression given by this combination of just two letters is almost nondescript. However, what they stand for is a plethora of technological possibilities that seemed inconceivable only a few years ago. "AI" – or Artificial Intelligence – is no longer just a secret code among tech nerds. It has literally taken up residence in our living rooms.
Daimler, too, is looking into the opportunities that AI applications provide for its customers and employees. Interlinked navigation systems, smart voice assistants, and autonomous vehicles are just a few examples of the potential we can leverage through targeted analysis of vast quantities of data.
Another thing is clear as well. The basic requirement for widespread acceptance of Artificial Intelligence is trust. In the past, intelligent systems were repeatedly the target of criticism. Such examples included chatbots using offensive language and discriminating facial recognition algorithms.
These cases show that the use of Artificial Intelligence requires some clear guidelines. For that reason, Daimler – as the first of any automobile manufacturer – has established a set of principles according to which we develop and deploy AI. They are intended to strengthen confidence in our company, products and services – while at the same time providing employees with guidance for dealing with AI.
1st principle: responsible use
"We shape and use Artificial Intelligence responsibly. We realize the opportunities of Artificial Intelligence and harmonize its effects in line with our corporate values."
AI is not an end in itself. Whenever we use it, it must have genuine benefits. To ensure that, our AI systems first undergo a transparent process of evaluation. Will they make our products, services and processes faster and more efficient? What advantage does using the technology have for different target groups? Is our data set representative, and is the quality high enough to enhance fairness and inclusion? We do not allow a machine to answer these questions. Humans remain in the driver’s seat.
2nd principle: explainability
"We are committed to a high level of transparency, thereby promoting trust in Artificial Intelligence. To that end, we support explainable Artificial Intelligence."
In school, our teachers taught us a special rule for solving math problems: The correct answer is not the only thing that counts. How you arrive at that answer is just as important. A similar rule applies to computations performed with the aid of Artificial Intelligence. When a piece of software, for instance, analyzes huge quantities of data in order to make predictions about future market developments, it should be obvious what parameters it considered while performing the analysis. That way, we can explain any possible unintended bias – and eliminate it.
3rd principle: protection of privacy
"We respect privacy. We consider privacy protection from the design phase of Artificial Intelligence systems onwards. We support privacy-enhancing technologies."
Digital services have now become a permanent part of our daily lives. They would not work without supplying a certain amount of data. That makes it even more important to ensure that the privacy of our customers and employees remains protected. In our MBUX multimedia system, personal information is processed exclusively inside the vehicle. Wherever possible and reasonable, we use mock or anonymized data records when testing and training AI systems. Compliance with legal regulations is always mandatory when we do. Data protection is a quality feature for us.
4th principle: safety and reliability
"We develop and test our AI technologies consciously and according to state-of-the-art science and technology. We take adequate measures to develop safe and reliable Artificial Intelligence."
Safety and reliability are quintessential parts of our brands. Therefore, we apply our high standards of quality to AI applications as well. Conscientious training of intelligent systems, scientifically validated and accepted algorithms, and usage of the latest technologies serve as their foundation. In autonomous vehicles, reliable AI is even safety relevant – another reason for us to work with exacting precision and a critical mindset when developing new technologies.
History shows that new technological developments always arouse concerns. Today, we still chuckle at anecdotes about people calling the first railroads "the work of the Devil" and the fear that their speed could supposedly "muddle the brain." I am convinced that we will be able to take advantage of the opportunities presented by AI for specific applications – with the aid of these principles and another form of intelligence that remains indispensable: human intelligence.