Organizational Culture Readiness for AI
How leaders set purpose, build literacy, and keep people at the center
AI isn’t a side project. It’s becoming part of how work gets done. For mission-driven organizations, the mission of delivering programs and services to people who rely on them hasn’t changed. And not every method will change; much of service delivery may look the same. What’s different is that, in the right circumstances, new tools can help us assess needs, deliver more efficiently and effectively, and analyze data to improve programs and decisions.
Over the past several decades we’ve gone from typewriters to laptops and tablets. Card catalogs, encyclopedias, and microfiche grounded us in past records; the internet added archives and real-time feeds; AI now pulls it all together—often in seconds—and can point to likely scenarios in the present or future while we use judgment. When used with care, this progression widens what’s possible and asks our organizational culture to evolve alongside it.
Leadership owns the shift
If AI is going to help, leaders go first. They set the purpose for using it; for example, faster response times, better access for residents and clients, less administrative burden on staff, and be clear about boundaries. Leaders model the behavior by using approved tools, sharing what worked and what didn’t, and encouraging careful skepticism over hype. They resource the work like infrastructure: fund training, privacy and security reviews, and give people time to learn.
Leaders can’t hand this to a working group, IT, or an innovation team and call it done. Those teams are essential partners, but ownership lives with executive and program leadership because AI changes how decisions are made and how services are delivered. Only leaders can set the vision and tone, make tradeoffs across programs, and hold the line on accuracy, human review and judgment, privacy, equity, and accountability. If leadership doesn’t own it, the work fragments and the culture won’t change.
This isn’t an initiative; it’s infrastructure
Initiatives come and go. Infrastructure stays. Think about AI the way we brought computers and the internet into work: new tools that change how information is found, created, and shared, and that show up in records, communications, and workflows. Learn what it’s good for and teach people to use it safely, with training built into onboarding and professional development.
Treating AI as infrastructure also means creating real budget lines for training, licenses, evaluation, and setting clear requirements in procurement: privacy, security, accessibility, auditability, and exit paths. RFPs and contracts should spell out data ownership, audit logs, and how you will turn tools off if needed. Folding plain language AI guidance into everyday documents, from staff handbooks to volunteer materials.
Build foundational literacy for everyone, then branch by role
Everyone needs a baseline of AI knowledge, then depth by role.
Start with fundamentals. In plain language, cover what generative tools do, where bias and privacy risks show up, and when not to use AI.
Practice with approved tools. Show simple ways to structure prompts and inputs so teams can safely draft FAQs or translate plain language summaries, and program staff can turn meeting notes into accurate updates for stakeholders.
Build critical thinking. Teach people to question outputs, check sources, spot missing voices, and escalate concerns.
Practice creating with AI. Draft, summarize, classify, map processes, and design the human review steps that keep judgment where it belongs.
Role-specific emphasis: supervisors focus on decision rights, risk triage, and measurement; technical and data teams go deeper on evaluation and audit trails; executives build enough comfort to model use and stay strong on governance.
Shortcuts, originality, and accuracy
Every generation has its shortcut. We debated whether to rely on condensed study guides, then website trustworthiness and copying. Now we ask whether we can trust an AI-generated draft. The answer is yes, with human judgment, review and clear rules. Require a quick accuracy check for factual claims and keep a short list of sources. For public-facing work, add a brief note about the process used and keep simple internal logs for higher-impact items. Protect privacy at every step and make accessibility a habit, not an afterthought.
Real risks: deepfakes, misinformation, and what’s ahead
Alongside the benefits are legitimate, serious concerns about capability and misuse. Synthetic media can impersonate elected officials or nonprofit leaders, fabricate “evidence,” or distort public debate. Models can hallucinate, reflect bias, leak sensitive data, or be manipulated by malicious inputs. That’s why training on risk recognition and response is as important as training on productivity use cases. Teach teams how to spot manipulated images or audio, authenticate official content through verified channels, and use detection tools as aids paired with human review. Establish a simple incident path for suspected misinformation or model failure, including who to notify, how to pause distribution, how to correct the record, and when to involve legal or privacy.
Thoughts to carry forward
We’ve done this before. We moved from typewriters to tablets, from microfiche and encyclopedias to the internet, from searching across scattered sources to having organized material presented to us on the spot. Each shift asked us to rethink how we work, how we learn, and how we show our work. AI is the next shift. If executives lead it on purpose, give everyone a foundation, prepare for real risks, keep people at the center, the organizational culture will be ready.
AI Disclosure: I knew what I wanted to write about for this post, starting with a detailed set of thoughts that I wanted to incorporate. I asked ChatGPT-5 Thinking to turn the thoughts into a draft post based on the tone and outcomes I was looking for, then GPT-5 Thinking and I worked through a couple more iterations. I made additional edits making sure it was true to my voice. GPT also helped me create the icon for this post.
Note: This post offers general guidance to help organizations plan AI work. It is not legal advice. Your approach should reflect your mission, values, applicable policies and laws, labor agreements, procurement rules, privacy and security standards, accessibility needs, and public records obligations. Please work with your counsel and internal teams to tailor what is right for your organization.