AI literacy, governance, and authenticity in higher education is about whether institutions can set rules for AI use that still let assessment prove what it is meant to prove. The strongest signal in the source set is that student AI use is now normal enough that “work must be your own” is too vague to carry the burden alone. The practical challenge is to distinguish legitimate support from hidden assistance, then design tasks, disclosure rules, and verification steps that still support valid, fair decisions. The evidence points towards policy, task design, and learner education working together; detectors and prohibitions alone are weaker foundations.

This topic covers the way higher education institutions teach AI literacy, write policy around AI use, and preserve the authenticity of assessed work when AI is widely available. The underlying assessment issue is not simply whether AI is used, but whether the resulting work still supports a defensible claim about learner achievement.

The source set shows a growing mismatch between institutional expectations and learner behaviour. When students can use generative AI easily and privately, assessment policy has to do more than state a ban; it has to explain what counts as independent work, what support is allowed, and how the institution will verify the claim being made. SUNY’s move to add AI education to information literacy is a sign that many systems are now treating AI as part of baseline academic capability, which changes the assessment context whether institutions like it or not. There is also a deeper issue of purpose. Dan Sarofian-Butin’s Chronicle essay asks whether education itself is at risk of becoming too performative if the work submitted no longer shows the learning underneath. That is a strong prompt for assessment teams to ask what the institution is really certifying: polished output, independent reasoning, AI-supported performance, or some mix of the three.

- **AI literacy**: the ability to understand what AI can and cannot do, and to use it appropriately. - **Authenticity**: whether submitted work genuinely reflects the learner’s own performance. - **Disclosure**: the requirement to state when and how AI was used. - **Verification**: the evidence needed to trust that the intended outcome was achieved. - **Governance**: the rules, roles, and review routes that make AI policy operational. - **Purpose clarity**: the need to be explicit about what the assessment is meant to prove.

The strongest evidence points to a shared practical conclusion: AI literacy has to be paired with assessment design and policy, not left as a standalone awareness exercise. SUNY’s move shows that institutions are beginning to build AI understanding into the curriculum, while the BBC and Chronicle pieces show why that matters for honest and meaningful assessment. The higher-education survey evidence reinforces the same direction: students are already using AI widely, so institutions need explicit expectations rather than implied norms. There is also broad convergence that assessment needs clearer boundaries. The better question is not whether AI should be present somewhere in the learner journey, but where it is acceptable, where it must be disclosed, and where the evidence must remain independent. That is a governance question as much as a pedagogic one.

What remains unsettled is how institutions should respond when learner behaviour moves faster than policy. One view is that stronger rules and detection can protect integrity; another is that the task must be redesigned so AI use is either explicitly included or less likely to obscure the learner’s own judgement. The source set does not settle that debate, but it does suggest that a policy-only response is too thin for live assessment practice. Another open question is where the line should sit between AI competence and academic independence. If institutions teach students how to use AI critically, some forms of support may become normal and legitimate. The unresolved issue is what additional evidence is needed when the same tools can also produce highly polished work with little visible learner contribution.

- Policy language may be too vague to guide real decisions. - Learners may use AI in ways that are hard to see after submission. - Institutions may over-rely on detection or sanctions rather than task design. - AI literacy may be taught without updating assessment rules. - Authenticity may weaken if final artefacts carry too much weight on their own. - Students may be punished for unclear rules rather than clear misconduct.

A sensible decision framework is: 1. Define what the assessment is meant to prove. 2. State clearly whether AI is prohibited, permitted with disclosure, or built into the task. 3. Teach students what those boundaries mean in practice. 4. Add process evidence where the learner’s own thinking matters. 5. Use detection, if at all, as one weak signal within a wider integrity process. 6. Review whether the task still gives valid evidence of the intended outcome. This sequence is more defensible than relying on general warnings about AI misuse. It pushes institutions to connect AI literacy, policy, and assessment design instead of treating them as separate problems.

| Option | What it means | Main benefit | Main trade-off | |---|---|---|---| | Prohibit AI | AI use is not allowed in the assessed task | Clear boundary and simpler rule-setting | Hard to enforce if AI use is already normal | | Permit AI with disclosure | AI is allowed but must be declared and bounded | More realistic and transparent | Needs very clear marking and verification rules | | Integrate AI into the construct | AI use is part of what is being assessed | Matches some real-world practice | Harder to preserve comparability and independence | The source set points towards the second and third options becoming more important, because institutions are increasingly being forced to define what kind of AI use still counts as valid evidence.

A university notices that students are using AI for planning, drafting, and proofreading, but the module handbook only says “submit your own work”. Rather than tightening the wording alone, the course team rewrites the brief to require disclosure, a short process note, and a follow-up discussion on key decisions. That makes the learner’s judgement visible and gives staff a better basis for deciding whether the work is authentic.

- BBC News article on a student reflecting on using AI to cheat at university. - Inside Higher Ed report on SUNY adding AI education to information literacy. - Chronicle essay on the purpose of education in the age of AI. - HEPI Student Generative AI Survey 2026. - Elon University / AAC&U survey on generative AI in higher education.

The visible market signal is not mainly about standalone products here; it is about institutions, advisory bodies, and media increasingly treating AI literacy as a core educational capability. Where vendors do appear in adjacent sources, they often frame the topic as productivity, support, or workflow improvement. That is useful context, but it does not answer the assessment question of what evidence is needed before AI-supported work can be trusted as independent learner performance.

### Should universities teach AI literacy separately from assessment rules? They can, but the two need to be connected. If students learn how to use AI responsibly but assessment briefs stay vague, the institution still leaves authenticity unresolved. ### Is “use your own work” still enough guidance? Usually not on its own. The source set suggests institutions now need to say what AI use is allowed, what has to be disclosed, and what evidence still needs to come from the learner. ### Can AI literacy improve assessment integrity? Yes, if it helps students understand the boundary between support and substitution. But literacy only helps when assessment tasks and marking rules are updated to match. ### What is the main unresolved question? How much process evidence, disclosure, or redesign is enough to make AI-rich work a trustworthy basis for a grade or award.

Tim Burnett (Admin)

Test Community Network. "AI literacy, governance, and authenticity in higher education." TCN AI & Assessment Wiki. Last reviewed 2026-05-03. https://www.testcommunity.network/wiki/ai-literacy-governance-and-authenticity-in-higher-education.html

- BBC News article on a student using AI to cheat at university. - Inside Higher Ed report on SUNY adding AI education to information literacy. - Chronicle essay on the purpose of education in the age of AI. - HEPI Student Generative AI Survey 2026. - Elon University / AAC&U survey on generative AI in higher education.