The AI Readiness Checklist & Basics Kit
A plain, honest starting point for any organization that keeps hearing about AI and is not sure where to begin. Run the self-assessment, read the short primer, and hand the whole thing to your team, board, or members.
Written for small businesses, chambers of commerce, and workforce and government teams. No jargon, no sales pitch, no cost.
Part one
The AI Readiness Checklist
Work through these honestly with the people who would actually use AI. Tick what is already true and watch your score update. There are no wrong answers. A “not yet” is just a place to start, not a failing grade.
Your readiness score
Totally normal. Pick two or three items below and make those your first month.
Leadership and goals
0/5Before tools, get clear on why you are doing this and who owns it. AI without a goal becomes a toy that nobody keeps using.
Data and tools
0/5AI is only as useful as what you can give it to work with. You do not need perfect systems, but you should know what you have.
People and skills
0/5The technology rarely fails. Adoption does. Readiness here is mostly about people feeling allowed and able to try.
Security and policy
0/5You do not need a legal department. You do need a few simple, written ground rules so people know what is and is not okay.
First use cases
0/5Start narrow. One small, real win that people can see beats a grand plan that never ships.
How to read your score: this is a compass, not a grade. A low number just means there is easy ground to gain. Focus on your weakest category first, pick the two or three items that matter most, and handle those before you roll AI out to more people.
Part two
The AI Basics Kit
A short, plain-language primer. No hype, no doom. Just what AI assistants and agents actually are in business terms, what to do first, what to avoid, and a glossary so the words stop getting in the way.
What an AI assistant or agent actually is
An AI assistant: is a tool you talk to in plain language to get help with a task, like drafting an email, summarizing a long document, answering a question, or rewriting something. You give it instructions, it gives you a draft, and you stay in control of what you do with it. Think of it as a very fast, very tireless helper that needs a good brief and a final review.
An AI agent: is an assistant set up to carry out a small task on its own, with rules you set, such as sorting incoming messages, pulling together a weekly summary, or drafting replies for you to approve. The difference is that an agent takes a few steps for you instead of waiting for each instruction. You still decide what it is allowed to do and where a human has to sign off.
What it is not: It is not a person, it does not truly understand your business, and it can confidently state things that are wrong. It is a pattern-matching tool trained on huge amounts of text. Used with a clear task and a human check, it saves real time. Trusted blindly, it creates new work and risk. The whole point of getting ready is to capture the time savings without inheriting the risk.
The first things any organization should do
Pick one painful, repetitive task
Choose a single task that happens often, eats time, and would not cause harm if a first attempt was rough. Drafting routine replies, summarizing notes, or turning a rough outline into a clean document are good starting points.
Try a reputable general assistant on it
Use a well-known AI assistant for that one task for a week. Do not buy a new system yet. The goal is to learn what AI is good and bad at on your real work, with no commitment.
Write down three simple ground rules
Before more people join in, agree on what not to paste into outside tools, who reviews output before it goes out, and which tasks are off-limits for now. Keep it to a single page.
Let one curious person lead
Give one motivated person a little time to get good at it and to help everyone else. Adoption spreads through a trusted colleague far better than through a mandate.
Review honestly, then decide
After a couple of weeks, ask plainly: did this save time, and was the quality acceptable after review? If yes, expand to one more task. If no, adjust the task or the instructions before giving up on it.
Common mistakes to avoid
Starting with the tool instead of the problem
Buying software before you know the task it serves leads to unused subscriptions. Name the problem first, then find the simplest thing that helps.
Trusting the output without checking it
AI can sound certain and still be wrong. Anything going to a customer, the public, or an official record needs a human review until you have earned trust on that specific task.
Pasting sensitive information into any tool
Treat outside AI tools like a public space until you have confirmed otherwise. Keep regulated, confidential, or client-protected information out unless you know exactly how it is handled.
Rolling it out to everyone at once
A wide launch with no champion and no ground rules usually fizzles. Prove one small win first, then expand deliberately.
Framing it as a way to cut headcount
If staff believe the goal is to replace them, they will quietly resist. Position AI as a way to remove busywork so people can do the work that needs a human.
Expecting perfection on the first try
The first attempts will be uneven. The skill is in writing clearer instructions and reviewing well, and that improves quickly with a little practice.
A starter glossary, in plain terms
- Prompt
- The instruction or question you give an AI tool. A clearer, more specific prompt produces a more useful result, the same way a clearer brief gets better work from a person.
- Model
- The underlying AI system that produces the responses. Different models have different strengths, costs, and limits. You usually access one through an assistant or app rather than directly.
- Generative AI
- AI that creates new content, such as text, images, or summaries, rather than just sorting or searching existing information.
- Hallucination
- When an AI states something false as if it were true. It is a known limitation, not a glitch, which is why a human review matters on anything important.
- Agent
- An AI set up to carry out a small task across a few steps on its own, within rules you define, instead of waiting for each instruction.
- Training data
- The large body of text and other material an AI learned from. It shapes what the tool knows and where its blind spots are.
- Fine-tuning
- Adjusting a general model with your own examples so it fits a specific task better. Most organizations do not need this to get value early on.
- Human in the loop
- A setup where a person reviews or approves what the AI produces before it is used. The simplest and most reliable safeguard for important work.
Keep a copy
Get a clean copy to share with your team
Everything above is free to use right now. If it is useful, we will email you a clean copy and a short set of starter resources you can forward to your staff, board, or members.