Editor
Celine Low chevron_right
As artificial intelligence increasingly shapes how we create and consume information, Malaysia is taking a proactive stance to ensure ethical practices. Communications Minister Datuk Fahmi Fadzil recently announced that Malaysia will soon introduce Guidelines for Responsible AI Communication to outline transparent and ethical practices across media, public relations, and digital content creation.
Speaking in a video address at the Kuala Lumpur International PR Conference (KLIP 7), Minister Fahmi emphasised, "Even as machines evolve, one thing must never change: human ethics and judgment must lead."
What's Currently On The National Guidelines on AI Governance & Ethics (AIGE)?
Launched in September 2024 by the Ministry of Science, Technology and Innovation (MOSTI), the National Guidelines on AI Governance & Ethics (AIGE) provide a broad, non-legally binding framework to promote responsible AI development and deployment.
These guidelines are intended to support the implementation of Malaysia's National AI Roadmap 2021-2025 and aim to build trustworthiness in AI while managing its risks and maximising its benefits.
The AIGE applies to AI end-users, policy makers (in government, agencies, organisations), and developers and providers of AI technology.
At its core, the AIGE outlines seven fundamental AI principles that guide the practice of responsible AI in Malaysia, which are
-
Fairness: AI systems are free from bias and discrimination.
-
Reliability, Safety, and Control: AI systems must be robustly tested, perform consistently, and be safe, secure, and resistant to unauthorised interference.
-
Privacy and Security: For the protection of personal data, AI systems should collect data with informed consent and store it securely, in compliance with data protection standards.
-
Inclusiveness: AI must be accessible and beneficial to all segments of society to cater to diverse and marginalised groups.
-
Transparency: AI systems must be understandable. This means clear communication about how they operate, how decisions are made, and disclosure of their purpose and training data (including any inherent biases).
-
Accountability: Developers, operators, and those who implement AI are held responsible for the AI systems' performance and outcomes, with clear documentation of responsibilities.
-
Pursuit of Human Benefit and Happiness: The ultimate goal is for AI to enhance human well-being, contribute positively to society, and respect individual rights.
Currently, Malaysia does not have a specific law regulating the use of AI. While existing laws (like the Personal Data Protection Act 2010) may apply to certain aspects of AI, they weren't designed with AI's unique complexities in mind.
As AI becomes increasingly prevalent, its exploitation has led to a rise in crimes such as deepfakes, cyberbullying, and cyberstalking, posing significant challenges that current laws in Malaysia are not yet equipped to handle.
Minister Fahmi also noted that Malaysia's chairmanship of ASEAN this year calls for leadership built on vision, empathy, and ethics, not just authority, principles which should be extended to the country's approach to AI.
What Could Be Covered in the Next One?
While the initial announcement provides a clear direction, the specifics of the guidelines have not been unveiled. It could cover the following:
-
Mandatory disclosure of AI-generated content (e.g., using watermarks or clear disclaimers).
-
Protocols for accuracy and fact-checking AI-generated information to prevent misinformation.
-
Specific measures for mitigating bias in AI-generated communications.
-
Clarifying copyright and intellectual property issues related to AI-created content.
-
Enforcement mechanisms and consequences for non-compliance.
-
Requirements for training and education for professionals using AI tools.
It's unclear when the guidelines will be finalised or announced. These guidelines are expected to complement the broader National AI Roadmap.
Stay updated with ProductNation on here, Instagram & TikTok as well.
More news here: