A.I Regulation and Media

The lecture is week six of the AI and media course, focusing on AI regulation. It covers arguments for and against regulation, global AI governance, the EU AI Act, Cambodia's AI industry governance, and AI regulation in media and communications.

Some argue against early AI regulation, stating it hinders progress and innovation. They suggest that AI itself is neutral, and its impact depends on its use and users, highlighting the importance of ethics.

Despite ethical considerations, regulation is needed because voluntary ethics are hard to enforce. Examples of AI misuse include deepfake pornography and AI-generated scam images. Regulation aims to prevent disruption in business, data privacy, and government programs, addressing issues like information disorder, data breaches, intellectual property infringement, and risks to health, safety, and fundamental rights.

One approach is to amend existing laws, like Denmark's decision to grant copyright to individuals' features to combat deepfakes. However, this method is impractical for the broad scope of AI. A more comprehensive approach is new legislation, exemplified by the EU AI Act.

Adopted in 2024, the EU AI Act classifies AI systems by risk: minimal, limited, high, and unacceptable. Minimal risk AI (e.g., spam filters) is unregulated. Limited risk AI (e.g., chatbots, deepfakes) requires transparency. High-risk AI (e.g., in law enforcement) is heavily regulated. Unacceptable risk AI (e.g., social scoring, mass surveillance) is banned within the EU.

Many countries, including Cambodia, are developing AI policies. Cambodia is in the early stages of drafting AI policies, as seen in a global visualization of AI regulation. Three key documents guide Cambodia's AI regulation: 'AI Landscape in Cambodia,' 'Cambodia Artificial Intelligence Readiness Assessment Report,' and the 'Draft National Artificial Intelligence Strategy'.

Cambodia's AI policy framework identifies national focus areas for AI development and regulation, calling for national regulations and guidelines, attention to AI infrastructure, promotion of AI research and education, development of an AI ecosystem, and international collaboration. Key sectors for focus include service, manufacturing, mobility, agriculture, education, finance, trade, and tourism.

Cambodia plans to develop AI regulations through a multi-ministry committee. These regulations will harmonize existing laws, create new data protection laws for sensitive personal information, adopt a 'privacy by design' approach, and oversee the integration of ethics and privacy in AI development aligned with international best practices.

A UNESCO-adopted methodology reveals Cambodia is at an early stage of AI adoption. Data, computing resources, and AI talent are limited. While universities and ministries collaborate on AI research, a comprehensive technical risk assessment is lacking, and no specific AI law is yet in place, though a sub-decree covers AI supply.

The report highlights ongoing work on laws for data protection, cybersecurity, and AI ethics, but notes the absence of laws for AI procurement and information access. The draft national AI strategy is nearing finalization. Socially, there are efforts to close the STEM gender gap and improve e-government services. There's a need for AI ethics integration in digital health and cultural heritage, as well as public awareness campaigns for safe digital skills and child protection. Scientifically, AI research is growing, despite limited budget, infrastructure, and an urgent need for curriculum upgrades and teacher training. Economically, AI is being used for societal benefit, but job displacement risk exists, necessitating ecosystem data gathering and strategic sector prioritization. Technically, draft laws for cloud-first and data governance are in progress, with a need to enhance connectivity, speed, and data centers.

The Royal Government of Cambodia has mandated specific roles for ministries in AI: the Ministry of Post and Telecommunications (MPTC) leads AI backbone infrastructure and governance; the Ministry of Industry, Science, Technology and Innovation (MISTI) handles AI research and development; and the Ministry of Education, Youth and Sport (MoEYS) coordinates AI education. All ministries are tasked with promoting AI within their sectors.

The draft National AI Strategy aims to complement existing national measures, promote AI adoption, define strategic directions for key sectors, establish AI regulations, increase inter-ministerial synergy, and encourage stakeholder participation. Strategic priorities include human resource development, data infrastructure, promoting AI adoption for digital government, sectoral adoption (including information), ethical AI, and fostering research, development, and innovation.

The MPTC will submit the draft National AI Strategy for review. This will be followed by the preparation of a national AI governance framework, involving workshops, submissions to committees, public feedback, and final approval by year-end. Concurrent sectoral AI strategy promotions will also take place to raise awareness and understanding.

The national AI strategy includes a strategic direction for the information sector, focusing on information integrity, clear communication, quick answers from reliable sources, and public sentiment analysis. A flagship activity involves three key ministries (Information, Interior, MPTC) to prepare new or updated regulations to combat AI-generated or manipulated content abuse. Existing relevant documents include the 1995 Law on the Press (under revision) and the charter for professional journalism.

The media's role includes promoting AI research, innovation, education, and talent development through scholarship. It also involves raising public awareness on digital skills, cybersecurity, and AI, and promoting related policies. Media and communication professionals should actively participate in policy discussions to ensure AI governance covers the information industry comprehensively.

The lecture summarized arguments for and against AI regulation, examined global AI governance with a focus on the EU AI Act, and detailed Cambodia's AI governance, key documents, and future directions, emphasizing the media's role in this development.