The utilisation of artificial intelligence in companies and society is increasing rapidly. The rapid development also raises fears and uncertainties concerning, for example, the fairness of the decisions made by artificial intelligence or the reliability of algorithms. That is why the Artificial Intelligence Programme draws attention to the ethics of artificial intelligence by challenging companies to get involved in the consideration of ethical principles.
Digia has also formulated principles for the utilisation of artificial intelligence.
“Artificial intelligence can appear to be an area that is difficult to understand, one that is associated with fears and worries, and AI applications may seem like black boxes. The purpose of ethical principles is to ensure that the utilisation of artificial intelligence is morally okay and that people are better able to assess the effectiveness of the decisions made by artificial intelligence”, says Markku Väänänen, head of the AI development project from Digia.
Digia’s principles for the ethical utilisation of artificial intelligence are:
1. Openness and responsibility
We enable the assessment of the artificial intelligence we have developed by describing, together with our customers, what is the aim of the algorithms, what kind of material is collected and why.
2. Transparency
We are transparent in utilising artificial intelligence and enable the same for our customers. We support our customers in, among other things, determining the accuracy of the data on which artificial intelligence is based, tracing the grounds for decision-making and creating rules for the utilisation of artificial intelligence. It is good to understand, for example, when artificial intelligence makes the actual decision and when it only supports people in decision-making with recommendations.
3. Assessability of impacts
We support our customers in the responsible utilisation of artificial intelligence and highlight the impact of choices related to our work.
4. Respect for privacy
We safeguard the privacy and personal data protection of the individuals represented in the data at our disposal in accordance with the information protection principles of our customer companies, partners and Digia as well as in compliance with the law.
“Once the ethical principles have been defined, they provide a clear basis for us Digia employees as well as our customers, partners and other cooperating parties. We can focus on the benefits to be gained through artificial intelligence while also being certain that the experiments and solutions implemented are ethical”, Markku Väänänen says.
Digia’s principles for the ethical utilisation of artificial intelligence were published on 7 March at the Women in Tech breakfast event organised at Digia.