Can the shortage of IT personnel in the labour market be partly addressed by deploying generative AI? This question was at the heart of the Commission Debate on Government Digitalisation held on 23 March 2023. State Secretary Alexandra van Huffelen requested CIO-Rijk to perform a quick scan to assess the impact of generative AI on government IT staff — a now completed task.
Responsible use of AI
Based on the quick scan results, and guided by advice from the State Attorney and the Dutch Data Protection Authority (AP), the government has formulated a preliminary stance on the application of generative AI. This commitment is towards responsible AI usage, entailing:
- Adhering to existing laws and regulations
- Conducting a risk analysis for each AI use case, including a Data Protection Impact Assessment (DPIA) and an algorithm impact assessment. These identify the risks and risk mitigation measures
- Submitting DPIA findings for consultation to the Chief Information Officer (CIO) and the Data Protection Officer (DPO)
These principles are also applicable to the utilisation or ongoing development of open-source generative AI applications.
Preliminary stance
The cabinet plans further discussions with municipalities and other government bodies to establish a government-wide position on this matter. The preliminary stance is also relevant when an external party provides services or products to a government organisation and employs generative AI. Furthermore, this stance applies to the experimental use of generative AI in pilots or testbeds.
