A practical guide to AI adoption at Reach University — built around who we serve, how we teach, and where we're going.
"Turn jobs into degrees. Reach brings higher education into the workplace and inspires deep learning through inquiry, dialogue, collaboration, and on-the-job practice — freeing minds, building careers, and strengthening the workforce."
Reach University is in the middle of remarkable growth. Enrollment has grown more than 600% since 2021 and is expected to grow another 50% in the year ahead. We're moving into new fields, new states, and new partnerships — all while staying true to the high-touch, relationship-based approach that makes our model work.
That kind of growth is only possible if the people doing the work — our faculty, advisors, operations staff, and partners — have the right tools to do more without burning out. AI is one of those tools. Not because it replaces the human relationships at the heart of the Reach Method, but because it handles the time-consuming tasks that get in the way of those relationships.
Think about what our Professors of Practice carry: they're working educators by day and teaching for Reach in the evenings. Our candidates are working adults with maybe 10 to 15 hours a week for coursework. Our Candidate Success Advisors support hundreds of people at once. AI helps everyone on that list do more of the work that matters by reducing the friction of the work that doesn't.
Reach's FY26-28 Strategic Plan notes the need to "develop new programs that meet the changing workforce dynamics caused by AI and technological advancements" and explicitly includes data infrastructure and technology capacity as a strategic priority. This AI strategy supports that priority directly.
There is also a bigger picture. Reach's moonshot — 3 million Apprenticeship Degree starts nationwide by 2035 — cannot happen without technology as a lever. The National Center for the Apprenticeship Degree (NCAD) needs to support partner institutions efficiently. Reach Teachers College needs to grow from roughly 3,200 candidates today to 7,500 without adding staff at the same rate. AI is part of how we get there responsibly.
↑ Back to topReach's FY26-28 Strategic Plan is organized around five priorities. AI is not a sixth priority on top of those — it's a tool that serves each one. Here's how:
AI helps faculty personalize feedback, gives candidates study support between evening sessions, and helps advisors spot who needs help before they fall behind.
AI helps Reach grow enrollment without growing overhead at the same rate — a key to reaching financial sustainability at 7,500+ candidates.
NCAD can use AI to help partner institutions design programs faster, access resources more easily, and build capacity without needing Reach staff on-site.
Domain 3 AI tools — chatbots, interactive reports, employer portals — are a direct expression of Reach's innovation priority and help tell the Apprenticeship Degree story at scale.
AI also connects directly to Reach's investment in data infrastructure. The IR and IT teams are developing data systems that are designed to integrate and share information across the university. AI tools that connect to those systems — or that generate better data — strengthen that foundation rather than competing with it.
↑ Back to topEvery AI use at Reach falls into one of three areas. These aren't rigid categories — they overlap — but they help different teams understand which kinds of AI use are most relevant to their work.
This domain is about supporting the experience of our candidates and the work of our faculty — without replacing the human relationship that makes Reach different from other online programs. The Reach Method is built on inquiry, dialogue, and practice. AI should create more space for those things, not crowd them out.
AI works best here when it handles preparation and feedback tasks so that synchronous sessions remain focused on the conversation and mentorship that only humans can provide. Professors of Practice should feel less burdened, not replaced.
Reach's AI guidance policy requires that AI use in academic work be disclosed. If a candidate or faculty member uses AI to generate ideas, text, or other outputs, they must include an appendix describing how they used it and why. This protects the integrity of the Reach degree.
This domain is about reducing the administrative load that comes with rapid growth. Reach is going from roughly 110 employees to 200–250 by FY28. AI can help current teams do more with less friction — drafting communications, organizing data, preparing documents — so that headcount growth stays focused on the roles that require human judgment.
No candidate names, Social Security numbers, financial records, health information, or any other personally identifying information should ever be entered into an AI tool. Always remove identifying details before using AI to work with any data about candidates, staff, or partners.
This is the frontier — building tools that bring AI-powered interaction to the people Reach serves and works with: employer partners, prospective candidates, funders, NCAD partner institutions, and policymakers. These are not off-the-shelf AI tools. They're custom applications that Reach builds or configures with IT support.
Any Domain 3 project should begin with a ticket to the IT office. These tools need to be reviewed for security, connected to approved data systems, and designed carefully. Don't build first and ask later.
IT has structured its AI support around five distinct deployment categories — moving beyond a simple "approved tools" list to reflect how AI is already present across our systems and how that landscape will grow. Each category has different access requirements, cost models, and governance implications. Knowing which category a tool falls into tells you how to get started and who to go to for support.
| Deployment | Status | Best Used For | How to Access |
|---|---|---|---|
| 1. Google Gemini Standard Workspace AI |
Available — All Users | Daily productivity: emails, notes, Docs, Sheets, research — available to candidates, faculty, staff, and contractors | Your existing Reach Google account |
| 2. Claude for Teams / Enterprise Licensed, IT-managed |
Licensed · IT-Managed | Deep analysis, long documents, complex drafting, custom tool development, and Domain 3 projects | IT provisioning via SSO; department budget required ($25/user/month) |
| 3. SaaS AI Functionality Vendor-embedded AI |
Emerging · IT-Activated | AI enhancements within existing enterprise systems (Canvas Ignite AI, Element BoltAI, future SIS) | Available through current vendor platforms; IT manages activation |
| 4. Marketplace Apps Google, Zoom, and platform add-ons |
IT Approval Required | Add-on tools found in Google Workspace, Zoom, and similar platform marketplaces | Submit an IT ticket before use — IT reviews and approves |
| 5. Claude Code / Azure AI Foundry Large-scale solutions |
Aspirational · IT-Led | Infrastructure-level AI projects: custom employer portals, university-wide integrations | Must run through IT — security and infrastructure review required |
| ChatGPT | Phasing Out | Not recommended for institutional use. If you're currently using it, please see transition support below. | Separate account — outside our enterprise agreement |
Gemini is built into your existing Google Workspace and is available to everyone at Reach — candidates, faculty, staff, and contractors alike. When you're in Gmail, Google Docs, or Google Drive, Gemini is already there, helping you draft faster, summarize longer threads, and work through routine tasks. IT will focus initial training and support primarily on faculty use cases, but there's no restricted access — anyone with a Reach Google account can start exploring today.
The AI features built into Gmail, Docs, and other Workspace apps are only available to users who are 18 or older. This means Gemini is appropriate for all Reach staff and faculty. For any AI tools used directly with candidates, check with IT first to understand which features apply.
Some of Gemini's most advanced features currently work best in English and are limited to U.S. users. Given that Reach serves multilingual communities and is expanding into diverse geographies, be aware that certain features may not work as expected for all workflows.
Claude for Teams / Enterprise is a licensed, controlled deployment at $25 per user per month, managed entirely by IT. Unlike using Claude individually, this version is tied to Reach's single sign-on, runs through our security infrastructure, and includes connector tools that integrate with Google Workspace, Slack, and other IT-managed platforms — all connector approvals go through IT. If your team is interested and has the budget, those funds flow to IT, which handles licensing and billing centrally. IT is also rolling out additional access controls limiting use to Reach accounts and Reach-managed devices — allow some runway for that rollout.
Claude is built with strong safety guidelines, which is a good thing — but it sometimes declines requests that touch on sensitive subjects, even when the request is entirely appropriate. If this happens, try rephrasing your request or providing more context about why you're asking. This may come up in behavioral health, mental health, or research contexts.
Many of Reach's existing enterprise vendors are embedding AI directly into their platforms. Canvas's Ignite AI enhances the learning management system. Element 451's BoltAI adds intelligence to our enrollment and student engagement tools. As our student information system evolves, SIS-level AI may follow. You don't install these separately — they appear as enhanced features within systems your team already uses. IT manages which SaaS AI features get activated and reviews each for data privacy compliance before enabling them.
Google Workspace, Zoom, and similar platforms have large marketplaces of third-party AI tools that supplement core functionality. These vary significantly in quality, privacy standards, and security practices. IT manages all requests and approvals for marketplace apps. Before adding any marketplace tool to your workflow — even a free or widely recommended one — submit an IT ticket. This isn't gatekeeping; it's ensuring we don't introduce data privacy or security risks through the back door of an app store.
At the most sophisticated end of the spectrum are large-scale, infrastructure-level AI projects — custom employer portals, university-wide candidate advising tools, or deep integrations built with platforms like Claude Code or Azure AI Foundry. These projects are less common but represent significant capability. Any project of this type must run through IT from the start — not as gatekeeping, but to ensure the proper security, infrastructure, and data governance are in place, and to minimize technology sprawl across the university.
ChatGPT doesn't connect to our Google environment, doesn't use your Reach login, and isn't covered by our institutional data agreement. Any data entered into ChatGPT — including details about candidates, partners, or Reach's operations — is outside our control. The transition to our five-deployment framework is about security and consistency, not restricting what you can do. Gemini and Claude for Teams together cover everything ChatGPT can do, and quite a bit more.
If your team has built workflows around ChatGPT, the IT team will work with you to map those workflows to Gemini or Claude for Teams. Submit a ticket to schedule a transition session.
↑ Back to topReach is not taking a top-down approach to AI. We're not telling faculty how to teach or telling operations teams which tasks to automate. We believe the people closest to the work are best positioned to identify where AI helps and where it doesn't.
What Reach provides are the boundaries that keep AI use safe, ethical, and consistent with our values. Within those boundaries, every team has the freedom to explore.
List three to five tasks your team finds time-consuming, repetitive, or error-prone. For each one, ask: could AI reduce the time this takes, improve the quality, or make it easier to do consistently? That list is your starting point. Share it with IT via a ticket and we'll help from there.
The most important rule is also the simplest: never enter information that identifies a specific person into an AI tool. This includes candidate names, student ID numbers, Social Security numbers, health information, disciplinary records, financial details, and any other personally identifying data. Strip that information before using AI to work with data, and when in doubt, ask IT first.
AI tools sometimes produce confident-sounding answers that are wrong. They can misunderstand context, miss nuance, or generate information that sounds plausible but isn't accurate. Always review AI output before using it, especially for anything that will be shared externally or used to make decisions.
Per Reach's AI Guidance Policy, if anyone — faculty or candidate — uses AI to generate ideas, text, images, or other outputs for academic work, they must document it: which tool they used, the prompts they gave it, the relevant outputs, and a brief explanation of how and why they used it. This isn't punitive — it's about transparency and integrity.
Both Gemini and Claude, through our enterprise agreements, protect Reach's data from being used to train future AI models. But you should still treat AI tools with the same care you'd give any university system. Don't enter anything you wouldn't be comfortable sharing through official Reach channels.
Reach's approach to education is built on inquiry, dialogue, collaboration, and practice. AI is a useful tool — not a replacement for the tutorial conversation, the mentoring relationship, or the human connection that sets Reach apart. If an AI use case would weaken those things, it's worth reconsidering.
↑ Back to topThe goal is to use AI where it genuinely helps Reach turn more jobs into more degrees. Start with one task, one tool, one team. See what works. Share it. Build from there. That's how the Reach Model has always worked — and it's how this will too.