A Digital Literacy in Prisons Device: Teaching Real-World App Skills Without Ever Going Online

Ask any prison education team what's missing from resettlement preparation, and digital literacy comes up fast; usually followed by a frustrated shrug. Everyday independence now runs through apps: banking, healthcare, Universal Credit, travel. Someone released after years inside can be handed a smartphone and expected to navigate all of it immediately, despite having had no safe way to learn or practice any of it while they were in custody.

The obstacle isn't willingness to teach it. It's that secure environments are (for good reason) locked down against exactly the things a normal teaching device would need: internet access, wifi, mobile connectivity. Any tool built for this has to work as a fully offline appliance, with no compromise on that point, while still giving prisoners something genuinely hands-on to learn from.

That's the problem SherlockAI set out to solve.

The Problem: Digital Exclusion Has No Practice Ground in Custody

Digital literacy rarely appears on a resettlement checklist alongside housing, employment and family contact — but it underpins almost all of them. Booking a GP appointment, managing a benefits claim, checking a bank balance, buying a bus ticket: these are no longer optional skills, they're the default interface to ordinary life. For someone who has been out of that world for years, the gap isn't conceptual. It's procedural. Nobody has shown them where the buttons are, what a confirmation screen looks like, or what to do when a login fails. That kind of fluency is normally built up gradually, almost invisibly, over years of everyday use — time that a custodial sentence removes entirely.

Prison education staff we've spoken with consistently raise the same issue: they want to teach this, but there's essentially no secure equipment built for it. Off-the-shelf devices assume connectivity. Consumer kiosk tools assume light-touch consumer environments, not a secure estate. The result is a genuine, practical gap in resettlement provision — one that's rarely visible until you go looking for a solution and find there isn't one.

Our Approach: A Secure, Offline Digital Literacy in Prisons Device

We've been building a working answer: a tablet that behaves like a normal smartphone from the user's point of view, loaded with realistic (but entirely simulated) versions of the everyday apps people will need on release; a mock banking app, an NHS-style health app, a travel/journey planner, a Universal Credit-style benefits app etc, while being technically incapable of reaching a network, a mobile mast, or the wider internet at any point.

Critically, the device is never "an app on a normal Android tablet with the wifi turned off." It runs on LineageOS, an open-source, independently auditable build of Android that we configure and compile ourselves, stripped of anything unnecessary — no Google account layer, no unused radios enabled, no consumer bloat. On top of that, we use Android's own built-in device-management framework — Device Owner and Lock Task Mode — to lock the tablet permanently into the training environment. This isn't a launcher app hoping a user doesn't find an exit; it's the same class of mechanism used to secure retail kiosks, hospital displays and managed school devices, applied here to a resettlement teaching tool.

The training content itself is deliberately simple to build and maintain: each mock app is a self-contained set of local web pages (HTML, CSS and JavaScript), so adding a new app or updating an existing one is a matter of copying a file across during an authorised maintenance step — not commissioning new software. That matters for a project like this, because the point isn't to freeze a fixed set of four apps in place; it's to build something a prison education team can keep shaping over time.

Built for Cost-Effective, Real-World Deployment

We're a not-for-profit CIC, and cost has shaped this project from day one — both because that's who we are, and because we know full well that secure-estate budgets don't stretch to bespoke enterprise procurement.

Two things keep this genuinely affordable:

• Reconditioned hardware. Through our parallel work reconditioning old laptops and tablets for reuse, we source low-cost, previously-used tablets rather than specifying new enterprise devices. A device that would otherwise be heading for landfill becomes secure teaching hardware instead — which also means the model scales cheaply: more devices simply means more reconditioned units, not a new procurement cycle.

• Open-source foundations. LineageOS, Android's own device-management APIs, and browser-based training content are all free, mature, and don't lock us into a vendor. There's no recurring licence fee sitting underneath this solution.

The result is a device that costs a fraction of a bespoke secure-tablet programme, while still being built on the same security primitives a formal prison security review would expect to see.

Co-Designing With Prison Education Teams

We don't think a not-for-profit sitting outside the wire should decide, alone, what a resettlement teaching tool should contain. The mock apps we've built so far are a starting point, not a finished product, and the version worth deploying at scale is the one shaped directly by the people who'll actually use it in a secure facility classroom.

We're actively looking to co-design this with prison education staff: which apps matter most, what functionality within them best mirrors the real friction people will hit on release, and what a security team would need to see before signing off on a device like this for real use. If you work in prison education, secure-estate IT, or offender resettlement and this is a gap you've been running into, we'd like to hear from you.

Get in Touch

If your prison, education provider, or resettlement organisation is interested in trialling a digital literacy in prisons device, or you simply want to compare notes on the same problem, get in touch with us. We're happy to share a demo, talk through the security architecture in as much or as little technical depth as useful, or just have a conversation about whether this fits what you're already trying to solve.