Linking Artificial Intelligence Principles
Article 5: Secure safe and controllable.
Enhance the measurability of ethical principles such as security and controllability, transparency and explainability, privacy protection, and diversity and inclusiveness; and simultaneously build corresponding assessment capabilities.
The highest principle of AI is safety and controllability.
Continuous efforts should be made to improve the maturity, robustness, reliability, and controllability of AI systems, so as to ensure the security for the data, the safety and security for the AI system itself, and the safety for the external environment where the AI system deploys.
1.2 Safety and controllability
Furthermore, the development of autonomous AI systems must have safeguards to ensure controllability of the AI system by humans, tailored to the specific context in which a particular system operates.
As specialists, members of the JSAI shall recognize the need for AI to be safe and acknowledge their responsibility in keeping AI under control.
In the development and use of AI, members of the JSAI will always pay attention to safety, controllability, and required confidentiality while ensuring that users of AI are provided appropriate and sufficient information.
with consideration of the diversity of AI systems so that: (a) the benefits of AI systems should increase through the sound progress of AI networking; and that (b) multiple developers’ efforts to control the risks should be coordinated well and operate effectively.
Principle of controllability
Developers should pay attention to the controllability of AI systems.
In order to assess the risks related to the controllability of AI systems, it is encouraged that developers make efforts to conduct verification and validation in advance.
In addition, in order to ensure the controllability of AI systems, it is encouraged that developers pay attention to whether the supervision (such as monitoring or warnings) and countermeasures (such as system shutdown, cut off from networks, or repairs) by humans or other trustworthy AI systems are effective, to the extent possible in light of the characteristics of the technologies to be adopted.
Verification and validation are methods for evaluating and controlling risks in advance.
● To make efforts to conduct verification and validation in advance in order to assess and control the risks related to the security of AI systems.
How should AI research and its applications, at the institutional level, be controlled?
In order to promote the healthy development of the new generation of AI, better balance between development and governance, ensure the safety, reliability and controllability of AI, support the economic, social, and environmental pillars of the UN sustainable development goals, and to jointly build a human community with a shared future, all stakeholders concerned with AI development should observe the following principles:
Safety and controllability
The transparency, interpretability, reliability, and controllability of AI systems should be improved continuously to make the systems more traceable, trustworthy, and easier to audit and monitor.
AI systems will be safe, secure and controllable by humans
AI systems should be verifiably secure and controllable throughout their operational lifetime, to the extent permitted by technology
The ARCC (Available, Reliable, Comprehensible, and controllable) for AI
Principle 4: AI should be controllable
Ensure the benefits substantially outweigh the controllable risks, and take appropriate measures to safeguard against the risks
AI systems must be secure, applicable and controllable.