Can artificial intelligence independently draft laws, conduct legal expertise, and predict the legal and socio-economic consequences? Experts are debating this question. Sholpan Saimova, head of the Center for Public Legislation at the Institute of Parliamentarism under the Administration of the President of the Republic of Kazakhstan and a PhD in Law, shares her views, reports infohub.kz.
In his State of the Nation address on September 1, 2023, President Kassym-Jomart Tokayev emphasized the strategic importance of digitalization and set a goal for Kazakhstan to become a leading IT country by adopting artificial intelligence technologies. This initiative was followed by the Concept for the Development of Artificial Intelligence for 2024-2029, approved by the government. The concept includes launching a supercomputer by 2025 and increasing the share of government services using AI to 20% by 2029. Work is also underway to digitalize parliamentary activities, including business process reengineering and the creation of a unified E-Parlament platform integrated with state information systems.
Experts estimate that by 2030, AI will contribute up to 7% of global GDP, and by 2027, the global AI market capitalization will exceed $400 billion. The use of AI in legislative activities is a global trend. For instance, Brazil passed a law entirely drafted by artificial intelligence, sparking debates about its role in lawmaking.
The Parliament of Kazakhstan performs three key functions: representative, oversight, and legislative. Digital efforts have so far focused mainly on the representative function, but integrating AI into the legislative process holds significant potential. The Institute of Parliamentarism is exploring three tools: using AI for legal expertise, developing predictive analytics based on data, and automating the creation of legislative texts. In 2020, the Anti-Corruption Agency allocated 2.4 billion tenge (without VAT) for scientific anti-corruption expertise of draft regulatory acts, and Al-Farabi Kazakh National University received 1.5 billion tenge. An example of the inefficiency of the old approach is the work on the draft Water Code, which revealed about 70 conceptual issues, a lack of follow-up on bylaws, and contradictions with constitutional norms. The Water Code took two years to develop.
To implement AI in the legislative process, a single operator is needed to develop tools and train specialists. The Institute of Parliamentarism is ready to act as such an operator. However, there are risks: cybersecurity, technical limitations in accessing government systems, and an underdeveloped linguistic field. It is important to focus on a 'quick wins' strategy — adopting conceptual decisions aimed at rapid implementation and tangible results. AI can transform the legislative process, improve accuracy, and reduce the time needed to draft bills, allowing the legal system to adapt quickly to changes.