Ethical black hole when the welfare digitized

the Researchers, Anne Kaun and Julia Velkova question whether Sweden is ready to let the machines make the decisions, within public welfare services. Swedish SSR parts the authors ' description of the problem, but believes that a clearer algoritmpolitik with sharper principles to meet the digitaliseringsutveckling that we face.

Decisions regarding the right to compensation within the welfare system will have far-reaching consequences for the citizens. Therefore, these decisions are always shaped on the basis of established principles of transparency, dialogue, transparency, the rule of law, and appeal. The principle of public access is a fundamental part. The same principles must apply if a machine/algorithm is the basis for decision-making.

There is a great potential in the algorithms and automated decision-making can perform routine tasks and relieve the employee, who then has the increased space to do what they are best at. This may involve contact with the public, provide support and help, and make complex judgments and to represent the authorities in the decisions that will help citizens to's and self-empowerment.

The principles on which the article presents, however, is far from sufficient, and it is difficult to see how they would have prevented, for example, datasystemfelet with the employment services as described in the article.

the Robots and algoritmernas force can be taken to be without the conditions for citizens and employees deteriorate. But then, a number of requirements must be met:

• Every decision must be linked to a responsible officer and orders should be followed and reviewed on the basis of which decision bases are and what information has the algorithm been programmed with.

• Where necessary, an individual needs assessment to complement the decision.

• Clear information for citizens about how decision-making goes to, what are the principles which govern and how to appeal.

• Algoritmernas codes shall be a public document and be able to be requested.

• A code of ethics should be drafted to ensure the work with algorithms.

• A algoritmombudsman up who has the power to get out of the codes that are the input for the algorithms and that can be a function of the citizens can turn to for support and information.