Bottom-Up AI: How Shaw University's Grassroots Strategy is Redefining Adoption

Opening Narrative

Ninety days ago, Shaw University began shifting its AI strategy from top-down automation to bottom-up empowerment. We believed real adoption would come not from executive mandates but from faculty and staff experimenting, testing, and shaping the tools for their own work.

That belief is holding. AI isn't just a side project anymore; it's becoming an integral part of how people work here.

The Approach: Small Starts, Real Wins

We avoided the typical "launch big" trap. Instead, we focused on exposure and ownership:

• Start small. Target early adopters first, the naturally curious.

• Keep it practical. Every session had to produce one real, ten-minute win.

• Track what matters. Measure progress by how ideas spread, not just attendance.

Behind the scenes, we built a lightweight, secure MCP layer, allowing one department to create something that works, and others can easily replicate it without needing a new system or committee.

The Results: Momentum That Sustains Itself

• 14% of staff have completed hands-on chatbot training.

• 6% of faculty have completed LearnWise AI training.

• By semester's end, participation is projected to rise another 8-11%, with faculty adoption expanding steadily.

Once one person in a department achieves a small win, others follow quickly. AI doesn't spread through memos; it spreads through proof.

What's Next

The next phase targets faculty workflows and course planning. We're moving from attendance tracking to usage tracking, measuring who's still applying AI tools two months later.

By December, our goal is clear:

• Half of the workforce through foundational AI training.

• Four or more departments actively using Copilot in weekly operations.

Lessons Learned for Other Institutions

If you're leading AI adoption in higher education, here's what's worked for us:

1. Start with volunteers, not mandates. Momentum grows faster when curiosity leads.

2. Measure stories, not just statistics. Reuse and ripple effects matter more than raw training counts.

3. Build infrastructure early. Even small wins need a path to travel, or they'll die in isolation.

Bottom-up AI isn't a loss of control; it's shared ownership. When people see the relevance, adoption follows naturally.

Closing Reflection

At Shaw, AI is moving from an initiative to an instinct. The most promising shift isn't technological:it's cultural.

We'll share our 180-day outcomes in January, focusing on who not only trained, but transformed how they work.