Please enable JavaScript.
Coggle requires JavaScript to display documents.
Future of AI safety in Thailand - Coggle Diagram
Future of AI safety in Thailand
Draft Decree
Categorizing AI into different awareness, following the EU AI Act
Unacceptable Risk (Prohibited AI)
Example
Social scoring system
AI that exploits people's vulnerbility
Obligation
Prohibit all use
Exception depending case by case basis
High Risk
Example
Chatbots
Deepfake
Content generated by AI
Obligation
Transparency Obligation
Users must be aware that they are interacting with AI
Content generated by AI must be labeled AI generated
Limited Risk
Example
Autonomous driving
AI in legal interpretation
Credit scoring
Obligation
Must be registered with regulators
Management measures
Risk control
Draft Act
Registration and consult with ETDA (Electronic Transactions Development Agency)
Shared Testing Sandbox for AI
Data sharing with the agency
Algorithm standard with equality
Standard Contractual Terms for AI Service provider
Obligation and responsibilities of a service provider.
Warranty of service standard
Risk Assesment & Strict Supervision
AI Compensation Fund
Confidentiallity of AI system data
The current flaw of AI safety in Thailand
Producing products without human involvement
The current Thai tort laws state that if there is a flaw that is harmful to the buyers, the controller of the production machine should take all responsibility
But what if the product has zero involvement from humans and the product is produced by AI
FAQ
What is the current state of law for AI safety?
What is the progress of developing AI safety in Thailand?
What is the difference between a Draft Decree and a Draft Act?
Which is preferable for SMEs using AI?
What is preferable for AI safety for the users?