5. The Principle of Explicability: “Operate transparently”
Published by: The European Commission’s High-Level Expert Group on Artificial Intelligence in Draft Ethics Guidelines for Trustworthy AI
Transparency is key to building and maintaining citizen’s trust in the developers of AI systems and AI systems themselves. Both technological and business model transparency matter from an ethical standpoint. Technological transparency implies that AI systems be auditable, comprehensible and intelligible by human beings at varying levels of comprehension and expertise. Business model transparency means that human beings are knowingly informed of the intention of developers and technology implementers of AI systems.
Explicability is a precondition for achieving informed consent from individuals interacting with AI systems and in order to ensure that the principle of explicability and non maleficence are achieved the requirement of informed consent should be sought. Explicability also requires accountability measures be put in place. Individuals and groups may request evidence of the baseline parameters and instructions given as inputs for AI decision making (the discovery or prediction sought by an AI system or the factors involved in the discovery or prediction made) by the organisations and developers of an AI system, the technology implementers, or another party in the supply chain.
3. Principle of controllability
Published by: Ministry of Internal Affairs and Communications (MIC), the Government of Japan in AI R&D Principles
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. One of the conceivable methods of risk assessment is to conduct experiments in a closed space such as in a laboratory or a sandbox in which security is ensured, at a stage before the practical application in society.
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. Generally, the former is used for confirming formal consistency, while the latter is used for confirming substantial validity. (See, e.g., The Future of Life Institute (FLI), Research Priorities for Robust and Beneficial Artificial Intelligence (2015)).
Examples of what to see in the risk assessment are risks of reward hacking in which AI systems formally achieve the goals assigned but substantially do not meet the developer's intents, and risks that AI systems work in ways that the developers have not intended due to the changes of their outputs and programs in the process of the utilization with their learning, etc. For reward hacking, see, e.g., Dario Amodei, Chris Olah, Jacob Steinhardt, Paul Christiano, John Schulman & Dan Mané, Concrete Problems in AI Safety, arXiv: 1606.06565 [cs.AI] (2016).
9. Principle of transparency
Published by: Ministry of Internal Affairs and Communications (MIC), the Government of Japan in Draft AI Utilization Principles
AI service providers and business users should pay attention to the verifiability of inputs outputs of AI systems or AI services and the explainability of their judgments.
Note: This principle is not intended to ask for the disclosure of algorithm, source code, or learning data. In interpreting this principle, privacy of individuals and trade secrets of enterprises are also taken into account.
[Main points to discuss]
A) Recording and preserving the inputs outputs of AI
In order to ensure the verifiability of the input and output of AI, AI service providers and business users may be expected to record and preserve the inputs and outputs.
In light of the characteristics of the technologies to be used and their usage, in what cases and to what extent are the inputs and outputs expected to be recorded and preserved? For example, in the case of using AI in fields where AI systems might harm the life, body, or property, such as the field of autonomous driving, the inputs and outputs of AI may be expected to be recorded and preserved to the extent whch is necessary for investigating the causes of accidents and preventing the recurrence of such accidents.
B) Ensuring explainability
AI service providers and business users may be expected to ensure explainability on the judgments of AI. In light of the characteristics of the technologies to be used and their usage, in what cases and to what extent is explainability expected to be ensured? Especially in the case of using AI in fields where the judgments of AI might have significant influences on individual rights and interests, such as the fields of medical care, personnel evaluation and recruitment and financing, explainability on the judgments of AI may be expected to be ensured. (For example, we have to pay attention to the current situation where deep learning has high prediction accuracy, but it is difficult to explain its judgment.)
We will make AI systems accountable
1. Accountability for the outcomes of an AI system lies not with the system itself but is apportioned between those who design, develop and deploy it
2. Developers should make efforts to mitigate the risks inherent in the systems they design
3. AI systems should have built in appeals procedures whereby users can challenge significant decisions
4. AI systems should be developed by diverse teams which include experts in the area in which the system will be deployed
1. Demand That AI Systems Are Transparent
A transparent artificial intelligence system is one in which it is possible to discover how, and why, the system made a decision, or in the case of a robot, acted the way it did.
A. We stress that open source code is neither necessary nor sufficient for transparency – clarity cannot be obfuscated by complexity.
B. For users, transparency is important because it builds trust in, and understanding of, the system, by providing a simple way for the user to understand what the system is doing and why.
C. For validation and certification of an AI system, transparency is important because it exposes the system’s processes for scrutiny.
D. If accidents occur, the AI will need to be transparent and accountable to an accident investigator, so the internal process that led to the accident can be understood.
E. Workers must have the right to demand transparency in the decisions and outcomes of AI systems as well as the underlying algorithms (see principle 4 below). This includes the right to appeal decisions made by AI algorithms, and having it reviewed by a human being.
F. Workers must be consulted on AI systems’ implementation, development and deployment.
G. Following an accident, judges, juries, lawyers, and expert witnesses involved in the trial process require transparency and accountability to inform evidence and decision making.
The principle of transparency is a prerequisite for ascertaining that the remaining principles are observed.
See Principle 2 below for operational solution.