As AI increasingly changes the nature of work, workers, customers and vendors need to have information about how AI systems operate so that they can understand how decisions are made. Their involvement will help to identify potential bias, errors and unintended outcomes. Transparency is not necessarily nor only a question of open source code. While in some circumstances open source code will be helpful, what is more important are clear, complete and testable explanations of what the system is doing and why. Intellectual property, and sometimes even cyber security, is rewarded by a lack of transparency. Innovation generally, including in algorithms, is a value that should be encouraged. How, then, are these competing values to be balanced? One possibility is to require algorithmic verifiability rather than full algorithmic disclosure. Algorithmic verifiability would require companies to disclose not the actual code driving the algorithm but information allowing the effect of their algorithms to be independently assessed. In the absence of transparency regarding their algorithms’ purpose and actual effect, it is impossible to ensure that competition, labour, workplace safety, privacy and liability laws are being upheld. When accidents occur, the AI and related data will need to be transparent and accountable to an accident investigator, so that the process that led to the accident can be understood.
Published by: The European Commission’s High-Level Expert Group on Artificial Intelligence in Draft Ethics Guidelines for Trustworthy AI
Trustworthy AI requires that algorithms are secure, reliable as well as robust enough to deal with errors or inconsistencies during the design, development, execution, deployment and use phase of the AI system, and to adequately cope with erroneous outcomes.
Reliability & Reproducibility. Trustworthiness requires that the accuracy of results can be confirmed and reproduced by independent evaluation. However, the complexity, non determinism and opacity of many AI systems, together with sensitivity to training model building conditions, can make it difficult to reproduce results. Currently there is an increased awareness within the AI research community that reproducibility is a critical requirement in the field. Reproducibility is essential to guarantee that results are consistent across different situations, computational frameworks and input data. The lack of reproducibility can lead to unintended discrimination in AI decisions.
Accuracy. Accuracy pertains to an AI’s confidence and ability to correctly classify information into the correct categories, or its ability to make correct predictions, recommendations, or decisions based on data or models. An explicit and well formed development and evaluation process can support, mitigate and correct unintended risks.
Resilience to Attack. AI systems, like all software systems, can include vulnerabilities that can allow them to be exploited by adversaries. Hacking is an important case of intentional harm, by which the system will purposefully follow a different course of action than its original purpose. If an AI system is attacked, the data as well as system behaviour can be changed, leading the system to make different decisions, or causing the system to shut down altogether. Systems and or data can also become corrupted, by malicious intention or by exposure to unexpected situations. Poor governance, by which it becomes possible to intentionally or unintentionally tamper with the data, or grant access to the algorithms to unauthorised entities, can also result in discrimination, erroneous decisions, or even physical harm.
Fall back plan. A secure AI has safeguards that enable a fall back plan in case of problems with the AI system. In some cases this can mean that the AI system switches from statistical to rule based procedure, in other cases it means that the system asks for a human operator before continuing the action.
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).
2. Transparent and explainable AI
We will be explicit about the kind of personal and or non personal data the AI systems uses as well as about the purpose the data is used for. When people directly interact with an AI system, we will be transparent to the users that this is the case.
When AI systems take, or support, decisions we take the technical and organizational measures required to guarantee a level of understanding adequate to the application area. In any case, if the decisions significantly affect people's lives, we will ensure we understand the logic behind the conclusions. This will also apply when we use third party technology.
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.