Fingrid’s ethical principles for artificial intelligence
We commit to develop and use artificial intelligence according to Fingrid’s values and operating principles. When utilising artificial intelligence, we also take care of people, the environmental impacts of our operations and good governance.
When referring to artificial intelligence, we primarily mean an application that learns to make decisions by itself and can adapt to different situations without human guidance. Our principles for artificial intelligence also apply to simpler applications when similar ethical questions are associated with them.
In practice, ethical use of artificial intelligence means eg. the following for us:
People are the core experts
We maintain a sufficient level of competence that allows us to utilise artificial intelligence responsibly for the benefit of Fingrid, our stakeholders and all of society. We ensure that people possess the core competence.
Clear responsibilities and monitoring
We understand the opportunities and risks of artificial intelligence. We clearly specify the responsibilities associated with artificial intelligence throughout all phases of the life cycle. People regularly monitor and supervise artificial intelligence to ensure that the activities are appropriate. As a company, we are responsible for the decisions made by our artificial intelligence.
Ensuring transparency and security
We describe our methods of using artificial intelligence in an open and understandable manner. We guarantee sufficient transparency and the traceability of decisions and the data on which those decisions are based. We always arrange appropriate data protection and data security practices.
For the good of people, fairly and humanely
We use artificial intelligence to help people by transferring routine work to a machine, we increase operational efficiency and improve quality. Our artificial intelligence must function equally for everyone. We have respect for people, and decisions that affect their legal status are always made by people.