You have access to the training data that was used to train an AI-based system. You can review this information and use it as a guideline when creating your tests. What type of characteristic is this?

Transparency is a quality characteristic of AI systems that refers to the ability to understand the system's internal workings, including its algorithms, data, and decision-making processes. Having access to the training data is a key aspect of transparency. It allows testers and stakeholders to investigate the data the model was built on, which helps in identifying potential sources of bias, understanding the model's learned patterns, and designing more effective test cases. This insight into the "why" behind the model's behavior is the essence of transparency.

A. Autonomy refers to the AI's ability to perform tasks independently of human control, not the ability for humans to inspect it.
B. Explorability is not a standard ISTQB term for this characteristic; it relates more generally to the ease of investigating a system's behavior.
D. Accessibility is a usability characteristic concerning the ease of use for people with diverse abilities and is unrelated to internal system inspection.

1. ISTQB, "ISTQB Certified Tester AI Testing (CT-AI) Syllabus Version 1.0," 2021. Section 1.2.2, page 9, defines transparency as "the extent to which the inner workings of the AI-based system are visible" and explicitly links it to having access to training data and algorithms.

2. Adadi, A., & Berrada, M. (2018). Peeking inside the black-box: a survey on explainable artificial intelligence (XAI). IEEE access, 6, 52138-52160. Section II defines transparency as a core concept of XAI, enabling understanding of how a model works. DOI: https://doi.org/10.1109/ACCESS.2018.2870052