UK student survey identifies access and AI-support gaps, with important sampling limits

Student-success leaders may have licensed AI tools without knowing who can meaningfully access them. Include student services, disability support, library lending, learning technology and procurement. Ask which students encounter device or connectivity problems and whether guidance reaches them before assessment. A bounded engagement could map one course's participation barriers and test a support pathway. The value hypothesis is fewer avoidable access failures, to be measured locally. The UK survey supplies discovery topics rather than a forecast for a U.S. institution. Do not promise learning gains, equate satisfaction with achievement or attribute broad digital barriers to AI alone.

Fit an access review to the actual student journey from identity login to course content and help. Prerequisites include supported-device definitions, accessible test cases and a lawful minimal telemetry plan. Test low bandwidth, mobile layouts, keyboard navigation, assistive technology and session recovery. Map outages and authentication failures separately from model-response quality. Protect disability-related support information from unnecessary propagation into assistant prompts.

Proposed proof of value

complete representative journeys on agreed devices and document each unresolved barrier with an owner. The survey does not specify a hosting solution or validate a product, and a successful usability check still cannot establish educational effectiveness.

Student services should coordinate the intervention with IT, faculty and accessibility staff. Inventory available loan devices, support channels and course guidance; recruit students to test the journey and train frontline staff to route problems. Dependencies include accessible materials, loan capacity and consistent permitted-use messages.

Proposed acceptance criteria

every reported critical barrier receives an owner and resolution route, representative journeys pass and students can identify available help. Report unresolved cases and support effort before expanding licenses. These are proposed service gates. Risks include biased feedback, confusing missing responses with satisfied students, and excluding learners who cannot access the survey itself.