Bridging the AI Divide: Why Community-Led Literacy Beats Top-Down Deployment

The University of Baltimore’s Cali Center suggests that AI adoption is a human challenge rather than a technical one. By categorizing a population into five personas, Builders, Adapters, Questioners, Skeptics, and Guardians, the center shows that resistance to AI often comes from a lack of personal agency. Current sales pitches for AI often deepen public distrust, which creates friction for future innovation. Leaders should move away from top-down implementation and toward ecosystem building. This is a slower, labor-intensive process that prioritizes trust in non-traditional spaces. Those who do this groundwork now will build a literate community, while others will spend their resources managing public backlash.

The High Cost of the Pro-AI Sales Pitch

Most organizations treat AI deployment as a technical rollout: install the software, train the staff, and expect results. Jessica Stansbury, founder of the Cali Center, argues this ignores the reality of community trust. When organizations push AI without including stakeholders, they strip people of their autonomy. This creates a cycle of resentment. The skepticism seen in trade workers or at public events is not a rejection of technology, but a rejection of being excluded from the decision-making process.

I feel like some of that booing is because what you are telling me this is now what is happening but I wasn't part of that conversation, you are making decisions for me, but I wasn't included.

-- Jessica Stansbury

The result is clear: by failing to establish why the technology matters to the individual, organizations create a permanent barrier to adoption. When the question of why it matters remains unanswered, the system hardens its defenses.

Why Obvious Solutions Fail Without Guardian Buy-in

Systems thinking requires us to look at who protects the status quo. Stansbury identifies Guardians, those who prioritize the stability of their community, as the most important stakeholders. If an institution tries to roll out a high-impact AI project, such as real-time overdose intervention, without the support of these Guardians, the project will fail regardless of its technical quality.

The temptation for leaders is to bypass this friction with top-down mandates. However, this ignores the reality of the last mile of implementation. If the people on the ground, such as substance users or neighborhood residents, do not trust the source, the technology becomes a liability.

You get the right people, you get the labor guardians who are protecting or the guardians who are protecting that group of people to go in and sit with them. I bet you you could walk out with them wearing some wearable devices because they don't want to die.

-- Jessica Stansbury

The advantage belongs to those who do the work of translation. By moving the conversation out of boardrooms and into spaces like barbershops or community centers, institutions can turn skeptics into participants. This is an investment in social infrastructure that improves long-term adoption.

The Power of So What: Reframing Institutional Roles

The shift from keeper of knowledge to facilitator of knowledge is the most important change for leaders. When an expert views themselves as a gatekeeper, AI feels like a threat to their identity. When they view themselves as a facilitator, AI becomes a tool for augmenting human potential.

This distinction determines the success of AI integration. Facilitators use AI to handle tedious administrative work, like report generation, which frees up time for human-centric community engagement. The benefit is that the institution acts as a neutral anchor, providing the literacy necessary for others to make informed choices. The institution becomes a trusted entity that provides the what, why, how, and where, rather than just another vendor pushing a pitch.

Key Action Items

• Audit your persona map: Over the next quarter, identify which groups in your organization fall into the categories of Builders, Adapters, Questioners, Skeptics, or Guardians. Do not treat them as a monolith.
• Shift from Pro-AI to Informed Choice: Stop selling the tool and start answering the so what for each persona. If you cannot explain why a skeptic should care, you are not ready for deployment.
• Move the conversation to safe spaces: Stop relying on town halls or formal lectures. Over the next 6 to 12 months, move your engagement to where your stakeholders already congregate, such as community centers or trade groups.
• Adopt the facilitator mindset: If you are in a leadership role, identify the administrative tasks that AI can automate to give your team more time for human interaction. This pays off in 12 to 18 months by reducing burnout and increasing trust.
• Prioritize human-in-the-loop verification: Establish strict protocols for human verification of AI outputs. As Stansbury notes, it is not a magic device. Investing in this rigor now prevents the cost of correcting AI-generated errors later.