You critically evaluate AI tools, AI-generated content and your interaction with the tool. You apply your human judgment to refine AI-generated content to create new understanding.
| Capabilities | Evidence of achievement |
|---|---|
| 3.1 Understand the importance of having a human in the loop | Use AI and non-AI approaches (prior knowledge, experience, professional judgment, evidence-based approach) to inform decisions. |
| 3.2 Manage common issues with AI-generated content (i.e. hallucinations, reliability of training data, limitations) |
Appraise an AI tool based on a range of factors (training data used, temporal coverage, knowledge of specialised topic, geographic or language biases). Understand what kind of information has been used to train AI models. Interpret XAI tools to improve transparency, trust, and accountability in AI. |
| 3.3 Assess the usefulness and relevance of AI-generated outputs |
Assess the accuracy of AI outputs and consider the sources it draws from using evaluative methodology (e.g. timeliness, relevancy, authority, accuracy, and purpose). Independently verify the accuracy and relevance of AI-generated information. |
| 3.4 Reflect on the AI interaction and change your approach if needed |
Implement strategies to protect privacy and confidentiality (e.g. data minimisation, anonymisation, and General Data Protection Regulation (GDPR) compliance). Adapt or refine your prompt to achieve a better output or solve problems in your output. |
| 3.5 Reflect on any ethical, legal and societal issues with the tool |
Assess whether the tool respects human rights, privacy, and individual autonomy. Evaluate if the tool respects intellectual property rights, licensing restrictions, and confidentiality requirements. |