The authorities shall develop guidelines with the aim of promoting the effective, safe and ethical use of generative AI in public administration. The guidelines shall serve as guidance and support for the use of generative AI for employees of state authorities, municipalities and regions from a legal, security and ethical perspective. The guidelines will also serve as methodological support and provide guidance for the activities' assessments of whether and how generative AI should be used.
Name in original language
Regeringsuppdrag till Myndigheten för digital förvaltning och Integritetsskyddsmyndigheten att ta fram riktlinjer för användningen av generativ AI inom den offentliga förvaltningen
Initiative overview
The Agency for Digital Governance (DIGG) and the Swedish Authority for Privacy Protection (IMY) will develop guidelines to promote the efficient, safe and ethical use of generative AI in public administration. The guidelines will serve as a guide and support in the use of generative AI for employees at government agencies, municipalities and regions from legal, security and ethical perspectives. In order to increase the use of AI in public administration, it is also important that people in management functions have access to the support needed to use the technology. The guidelines will therefore also serve as a methodological support and, based on legal, security and ethical aspects, provide guidance for the operations' assessments of whether and, if so, how generative AI should be used. In this respect, the guidelines will also support public administration in taking advantage of generative AI in a way that fosters innovation and development. However, the guidelines developed within the framework of the assignment should not target the use of generative AI in education or research. If the authorities deem it appropriate, the guidelines can be supplemented with training material or a sample catalogue, to show the possibilities of using generative AI in public administration. The activities conducted in public administration are very broad, which means that generative AI is already used to varying degrees. Internal guidelines or recommendations for generative AI have been developed to varying degrees within different government agencies, which can contribute expertise and experience of the use of generative AI in public administration. The purposes of the use of generative AI in public administration differ between different activities. In addition, depending on their size, the businesses have different conditions for promoting the use of AI on their own. The authorities must therefore conduct a dialogue with the relevant public and private actors during the implementation of the assignment. The work on the guidelines will make use of already existing knowledge support, e.g. The trust model for AI managed by DIGG. The guidelines shall be based on current regulations and be based on the state's core values in order to create trust, transparency and build trust. In so far as the guidance relates to data protection, it is produced and decided by the IMY in its capacity as an independent data protection authority. However, DIGG will coordinate the authorities' work to develop the guidelines. The assignment will also contribute to increasing the level of knowledge about the opportunities and limitations of AI in public administration. During the implementation of the assignment, DIGG will offer competence-enhancing initiatives to smaller actors in public administration in order to promote the administration's appropriate and efficient use of AI. DIGG can offer smaller municipalities, regions and authorities support in how they can use AI.
Name of responsible organisation (in English)
The Swedish Government, the Swedish Agency for Digital Governance and the Swedish Authority for Privacy Protection
