AI learning support can improve access, practice, confidence, and feedback, but it can also change what later counts as independent assessment evidence. The key assessment question is where support ends and the evidence claim begins. New source notes in this run strengthen that boundary problem: self-directed experimentation, classroom AI, and personalised study tools may be helpful, but they also make preparation less visible and more variable across learners. The practical task for assessment teams is to define the boundary clearly enough that support does not become hidden assistance.

AI learning support refers to tools and practices that help learners study, practise, revise, generate questions, receive guided feedback, or experiment with ideas before they reach a formal assessment. In assessment terms, the issue is whether the support improves the learner’s ability to demonstrate the intended capability, or whether it changes the meaning of the capability being evidenced.

This topic matters because assessment does not begin at submission; it begins in the learner environment. If students use AI to build quizzes, summaries, study plans, or guided learning routines, they may arrive better prepared, but also with more of the cognitive work already offloaded. That can be positive for learning and negative for assessment if the final task is supposed to evidence independent thinking, drafting, or problem-solving. The more structured the AI support becomes, the more explicit the boundary between practice and evidence needs to be.

- **Preparation support**: AI that helps the learner study before assessment. - **Guided learning**: AI that takes the learner through steps rather than simply giving answers. - **Cognitive offloading**: shifting some mental work from the learner to the tool. - **Boundary setting**: making clear what is allowed in preparation and what is allowed in the assessed task. - **Assessment evidence**: the work or performance used to make the qualification claim.

The source set points towards a broad practical consensus that AI learning support is now normal enough to matter for assessment. Study mode, quizzes, summaries, and self-directed experimentation all suggest that learners will increasingly arrive with AI-shaped preparation habits. There is also a clear assessment implication: learning support can be legitimate and still be irrelevant or problematic in the final evidence claim. A tool that is fine for revision may be inappropriate inside the assessed task if it changes what the learner is meant to show.

The open question is where helpful scaffolding ends and hidden assistance begins. Self-directed experimentation and guided study features can improve confidence and engagement, but they can also make it harder to know how much independent thinking the learner has done. The evidence in this update is mainly practitioner and product framing, so the question for readers is how much support is compatible with the specific assessment construct. Another unresolved issue is comparability. If one learner uses heavily supported study tools and another does not, the final assessment may still be comparable if the test is strong enough — but that assumption needs checking, not guessing.

- Learners may over-rely on AI support and under-develop independent capability. - Preparation support may become hidden assistance if boundaries are vague. - Confidence may rise faster than mastery. - AI-supported revision may create uneven advantage across learners. - Assessment teams may assume the learner environment is neutral when it is not.

A useful decision framework is: 1. Define what the assessment is meant to evidence. 2. Map the AI support learners are likely to use before the assessment. 3. Decide which support is acceptable for practice and which is not acceptable in the final task. 4. Write the boundary into learner-facing guidance, examples, and mark schemes. 5. Check whether the assessment still requires enough unaided thinking to support the intended claim. 6. Revisit the boundary if the learning environment changes materially.

| Option | What it means | Main benefit | Main concern | |---|---|---|---| | Support for preparation only | AI is allowed in study, revision, and practice | Helps access and confidence | Can still change learner habits | | Limited support in selected tasks | AI is allowed only where the construct tolerates it | More realistic and flexible | Needs clear disclosure rules | | AI-integrated performance | AI use is part of the assessed capability | Aligns with AI-rich practice | Harder to compare and defend |

A student uses an AI study tool to turn notes into quizzes and summaries, then uses a guided study feature to work through a topic step by step. That is likely helpful for revision, but the assessment team still decides that the final exam must show independent reasoning without AI assistance. The policy is clear that preparation support is allowed, but the exam itself is the evidence point.

- Self-directed experimentation and personalised education. - Guided study support. - Study support, quizzes, and flashcards. - Note-to-quiz and summary study support.

The market is clearly moving towards AI study tools that promise better preparation, retention, and convenience. That is useful context, but it does not prove that these tools improve learning in ways that support assessment validity. Buyers and assessment teams should treat vendor claims as prompts to define boundaries, not as evidence that the boundary problem is solved.

### Is AI learning support the same as assessment support? No. It may help preparation, but it is not automatically acceptable inside the assessed task. ### Can study tools make assessment unfair? They can if some learners have much more effective access than others or if the assessment assumes a neutral preparation baseline. ### What is the main risk? That useful preparation becomes hidden assistance when the learner submits work or sits the test. ### What should assessment teams ask first? Ask what the assessment is meant to evidence and what AI support is likely to shape the learner before they reach it.

Tim Burnett (Admin)

Test Community Network. "AI assessment and learning support." TCN AI & Assessment Wiki. Last reviewed 2026-09-28. https://testcommunity.network/knowledge/ai-in-assessment/wiki/ai-assessment-and-learning-support.html

- Self-directed experimentation and personalised education. - Student AI problem-solving and human values in AI-driven environments. - Guided study support. - Study support, quizzes, and flashcards. - Note-to-quiz and summary study support.