AI for Small Business: The 4D Framework for Working Smarter (and Safer) with AI

Why AI for small business is a skill, not a subscription

With AI for small business, the companies getting real value aren’t the ones with the most tools. They’re the ones whose people know how to work with AI: what to hand over, how to ask, how to check the result and when to take responsibility for it.

That skill has a name: AI fluency. Researchers Rick Dakan and Joseph Feller, working with Anthropic, organized it into four competencies called the 4D framework: Delegation, Description, Discernment and Diligence. We walked through it in our webinar, Working Smarter with AI, and this article is the written version.

Whether you run a 15-person engineering firm in Golden or a nonprofit in Five Points, the 4Ds give your team a shared, practical way to use AI well. They also form the foundation of good AI governance, without a 40-page policy binder.

In this article:

• The three ways people work with AI
• The 4Ds, with real small-business examples
• How the 4Ds connect to shadow AI, AI policies and AI governance
• Answers to common questions about AI for small business

The three ways to work with AI

Every time someone on your team uses AI, they’re working in one of three modes: automation, augmentation or agency. Knowing which mode you’re in tells you how much attention the work needs.

Mode	What it means	Small-business example	Your role
Automation	AI completes a task you define	Turning meeting notes into a summary email	Define the task, review the output
Augmentation	You and AI think it through together	Brainstorming a bid strategy for a construction project	Bring the judgment; AI brings options
Agency	AI acts on its own, inside limits you set	An AI assistant that triages the support inbox	Set the boundaries, monitor the results

The natural path for AI for small business is to start with automation, grow into augmentation, and approach agency carefully. The more independence AI has, the more the 4Ds matter.

The 4D framework at a glance: four skills, one question each

The 4D framework boils AI fluency down to four skills. Each comes with one question to ask yourself every time you use AI.

Skill	What it means	The question to ask
Delegation	Decide what AI should do, and what stays with you	“Should AI do this, and which part?”
Description	Tell AI clearly what you want, and how	“Did I give it what it needs?”
Discernment	Judge what comes back with what you already know	“Is this right, and how do I know?”
Diligence	Own the result, and how AI was used to get there	“Would I put my name on this?”

If your team remembers nothing else, these four questions will carry most of the weight.

D1, Delegation: decide what to hand over

Delegation means deciding which parts of a job AI should do and which parts stay with you. The question isn’t “can AI do this?” It’s “which part should AI do, and which part stays with me?”

Ask yourself: “Should AI do this, and which part?”

Good delegation draws on three kinds of awareness:

• Problem awareness: knowing the goal, and what a good result looks like, before you start.
• Platform awareness: knowing what your AI tools are good at, where they fall short, and which tools are approved for your business data.
• Task delegation: splitting the work so people and AI each do what they do best.

Example: a client proposal. AI drafts the scope summary and cleans up the formatting. You set the price, choose the terms and own the relationship.

Platform awareness is where AI for small business most often goes wrong. When employees don’t know which tools are approved, they pick their own. That’s how shadow AI starts: company or client data pasted into personal accounts no one is managing. An AI readiness assessment is often the fastest way to learn which tools your team already uses.

D2, Description: ask for what you actually want

Description is how clearly you tell AI what you need. The gap between a useless answer and a useful one is almost always what went into the request.

Ask yourself: “Did I give it what it needs?”

A strong request covers three things:

• The product: what you want back, including the format, the length and the audience.
• The process: how to approach it, such as the steps to follow, the sources to use, or what to leave out.
• The performance: how AI should work with you, such as asking clarifying questions, pushing back on weak ideas, or keeping a formal tone.

Example: a law firm client update. “Summarize this case status” gets a generic paragraph. “Write a 150-word update for a non-lawyer client, in plain English, covering the hearing date and next steps, without legal advice” gets something you can actually send.

Tip: Describe the task the way you’d brief a sharp new hire on their first day. If they’d need more context, so does AI.

D3, Discernment: check the work like an expert

Discernment means evaluating what AI gives you before you rely on it. AI can sound completely confident and still be wrong. Your experience is the filter.

Ask yourself: “Is this right, and how do I know?”

Check three things:

• The product: Is it accurate, complete and right for the audience? Are the facts, figures and citations real?
• The process: Did AI follow your instructions, or did it skip steps and fill gaps with guesses?
• The performance: Is AI working the way you need it to, or should you change how you’re asking?

Example: a manufacturing safety procedure. AI can draft a lockout/tagout checklist in seconds. Only someone who knows your equipment can tell whether a step is missing, and a missing step matters.

This is where your team’s expertise becomes more valuable, not less. AI handles the first draft; your people make sure it’s right.

D4, Diligence: own what goes out the door

Diligence means owning the result, and how AI was used to get there. AI can do the work, but a person still stands behind it.

Ask yourself: “Would I put my name on this?”

Diligence shows up at three points:

• Creation: choosing tools and uses that are ethical, secure and appropriate for your data.
• Transparency: being honest with clients, colleagues and partners about when and how AI was used.
• Deployment: verifying the final work before it goes out, and owning it once it does.

Example: a healthcare practice. A front-desk team uses AI to draft patient reminder messages. Diligence means using a HIPAA-compliant AI tool, never pasting patient details into a personal chatbot, and reviewing each template before it’s sent.

Diligence is where AI governance starts

AI governance sounds like an enterprise concern. For a small business, it simply means agreeing on how your team uses AI, and writing it down. Without that agreement, shadow AI fills the gap.

A practical starting point:

1. An AI acceptable use policy that names approved tools, what data can and can’t go into them, and who reviews AI-assisted work. An AI policy template gets you most of the way there.
2. An approved tool list that matches your security and compliance needs, whether that’s HIPAA for healthcare, client confidentiality for law firms, or donor data for nonprofits.
3. Training on the 4Ds, so everyone shares the same habits and the same language.

Frequently asked questions about AI for small business

What is the 4D framework for AI?

The 4D framework is a model of AI fluency built on four competencies: Delegation, Description, Discernment and Diligence. It was developed by Rick Dakan and Joseph Feller with Anthropic to help people work with AI effectively, efficiently, ethically and safely.

How should a small business start using AI?

The best first step with AI for small business is a low-risk, repetitive task such as summarizing notes or drafting routine emails. Use an approved tool, describe the task clearly and review the output. Then expand to new uses once your team is comfortable.

What is shadow AI?

Shadow AI is the use of AI tools at work without the business knowing about them or approving them. The main risk is company or client data going into personal accounts that no one manages or secures.

Does a small business need an AI policy?

Yes. If your employees use AI, a short AI acceptable use policy protects your data and sets clear expectations. It should name approved tools, define what data is off-limits, and say who reviews AI-assisted work.

What is AI governance for a small business?

AI governance is the set of rules and habits that guide how your business uses AI. For most small businesses, that means an AI policy, an approved tool list and basic training, not a large compliance program.

What is an AI readiness assessment?

An AI readiness assessment reviews your current tools, data, security and team skills to show where AI can help and what needs to happen first. It gives you a clear, prioritized starting point.

What are the best AI tools for small business?

The best AI for small business is the tool that fits the systems you already use and protects your data. If your team works in Microsoft 365, start with Microsoft Copilot. Whatever you choose, use a business plan with admin controls that keeps your data out of model training.

Join us live: Working Smarter with AI

Want to see the 4Ds in action? Join our free virtual webinar, AI Simplified: Working Smarter with AI, on October 29 at 11:30 a.m. MT.

eCreek CTO Colter Strahan will walk through the 4D framework, share prompting techniques you can use the same day, and show how to keep shadow AI from putting your data at risk. Elan Jones, our Director of Business Development, will moderate and take your questions live. You’ll also leave with a reusable prompt template your whole team can use.

Register now to save your seat.

Put the 4Ds to work in your business

AI for small business doesn’t have to be complicated. The 4Ds give your team a common language: hand over the right work, ask clearly, check carefully and own the result.

As Your IT Department, eCreek helps Colorado businesses and nonprofits adopt AI with the same care we bring to cybersecurity and compliance. We help you choose the right tools, set sensible guardrails and train your people. And we’re local, so we’re never far away.

• Not sure where to start? Book an AI readiness assessment.
• Need guardrails now? Download our free AI policy template.
• Want the full walkthrough? Register for Working Smarter with AI on October 29.

The 4D framework comes from the AI Fluency Framework by Rick Dakan and Joseph Feller, developed with Anthropic.

Sources

• The AI Fluency Framework, Rick Dakan and Joseph Feller