The Senedd Research, on behalf of the Welsh Parliament, has released a detailed overview of Generative AI, examining the opportunities and risks it presents in the public sector. This report also highlights the current regulatory framework in Wales and other countries, as well as the challenges involved in implementing these technologies.
Opportunities and Risks of Generative AI in the Public Sector
Generative artificial intelligence (GenAI) refers to the use of AI for creating new content. The report outlines several potential benefits of GenAI for the public sector, including improvements in efficiency, quality, and accessibility of public services. Additionally, GenAI offers unique opportunities to promote and support the Welsh language.
However, the report cautions that the outputs of generative AI are highly dependent on the quality of the data used to train models. This can lead to issues such as bias, discrimination, mistakes, or disinformation. There are also challenges related to the availability of training data in languages like Welsh, which could limit AI capabilities. Beyond these technical limitations, the report identifies several broader risks, including a lack of regulation, potential impacts on jobs, privacy concerns, data protection issues, and copyright challenges.
Current Approach to Regulating Generative AI in Wales
The report notes that there is no specific holistic approach to legislating for generative AI in Wales. Current legislative tools rely on existing regulatory frameworks, which include domain-specific legislation and broader regulations that apply across multiple domains. Key areas such as the protection of personal information, intellectual property, and internet services are reserved matters, meaning that the Welsh Government and the Senedd do not have the power to make laws in these areas, leaving Wales subject to UK-wide legislation.
The UK Government published its AI regulation white paper in March 2023, advocating a ‘pro-innovation’ approach. This strategy does not propose new specific regulations or regulatory bodies but emphasizes that AI should be governed within the existing powers of regulators. The UK Government’s response in February 2024 indicated support from stakeholders and noted that regulators are already acting in accordance with the proposed framework.
International Approaches and Challenges
Other countries have adopted different approaches to AI regulation. For example, Brazil and China have developed prescriptive legislative frameworks, while the EU has been at the forefront with its AI Act. The report emphasizes that regulation needs to be flexible to accommodate the rapid evolution of generative AI technologies. Defining the impacts of generative AI can be challenging, making it difficult to identify who has been harmed and when. Furthermore, international collaboration is essential to address these challenges effectively and avoid displacing problems. The AI Safety Summit in November 2023 highlighted these collaborative efforts, although strategies still vary by country.
Read the Full Report
The full report by Senedd Research provides an in-depth analysis of the opportunities, risks, and regulatory challenges associated with generative AI in the public sector. It is a crucial resource for policymakers, industry leaders, and stakeholders interested in understanding and navigating the complex landscape of AI regulation.
To explore the detailed findings and recommendations, click through to read the full report here.
This comprehensive overview underscores the need for a balanced approach to harness the benefits of generative AI while mitigating its risks, ensuring that technological advancements contribute positively to society.