AI Tools for VAs
AI will not replace VAs, but VAs who use AI will out-earn those who do not. Learn to use AI to work faster, deliver more, and charge more. Complete all 6 modules to earn your AI Tools certificate.
Module 1: AI Foundations for VAs
- Explain what AI is good at and where it still needs a human check.
- Write a clear, structured prompt that gets a usable answer on the first try.
- Choose between ChatGPT, Claude, and Gemini for a given task.
- Set honest expectations with clients about how you use AI in your work.
Lesson 1.1 What AI Can and Cannot Do for a VA
AI is a fast assistant, not a replacement for your judgment. Knowing exactly where it helps and where it can quietly mislead you is the first skill every AI-enabled VA needs.
- Good at: drafting first versions of emails, posts, and documents fast.
- Good at: summarizing long text and extracting data into lists or tables.
- Good at: brainstorming options, rewriting for tone, and catching typos.
- Not good at: knowing facts specific to a client's business unless you give it to them.
- Not good at: guaranteeing accuracy, it can sound confident while being wrong.
- Not good at: replacing your judgment on anything sensitive, financial, or public facing.
Lesson 1.2 Writing a Prompt That Actually Works
Most weak AI output comes from a weak prompt. A strong prompt gives the AI four things every time: a role, a task, context, and a format.
- Role: tell the AI who it should act as, for example "you are an email copywriter for a skincare brand."
- Task: state exactly what you want done, one clear action.
- Context: give it the background it needs: the client's voice, the audience, and the real facts.
- Format: say exactly what the output should look like: length, bullet points, or tone.
- Weak prompt: "write an email about our sale."
- Strong prompt: "You are an email copywriter for a skincare brand. Write a 100 word email announcing our 20 percent off sale this weekend. Warm, friendly tone. End with a clear call to action button labeled Shop Now."
Lesson 1.3 The Main Tools and When to Use Each
ChatGPT, Claude, and Gemini all do similar things well, but each has a lane where it tends to shine.
ChatGPT
A broad, general purpose tool. Strong for brainstorming, casual writing, and quick answers. Widely used and easy to start with.
Claude
Strong for longer documents, careful writing, and following detailed instructions closely. A good choice when tone and nuance matter.
Gemini
Built into Google Workspace. Useful when you are already living in Gmail, Docs, and Sheets and want AI right where you work.
Tell a client which tool you used and why, in one plain sentence. Confidence about your process builds trust faster than hiding it.
Turning a vague ask into a usable draft
- The client says: "Can you help me with an email about our new product?"
- You ask two clarifying questions: what is the product, and what should the reader do next.
- You write a structured prompt: role (email copywriter for the brand), task (write a 150 word launch email), context (product details and audience), format (subject line, three short paragraphs, one call to action).
- You run it, then edit two lines that sound generic.
- You send the client a polished draft in minutes instead of starting from a blank page.
Build your AI foundation
Prompt framework
A reusable structure for any AI task. Fill in the blanks and reuse it every time.
You ask AI to check a competitor's pricing, and it gives you a confident, specific answer. What is your best next move before you use it?
Module 1 Quiz
What is AI generally best at for a VA's daily work?
Which part of a strong prompt tells the AI who it should act as?
A prompt says "write about our sale." What is missing that would make it stronger?
When should you double check something AI gave you before sending it to a client?
Module 2: Writing and Content with AI
- Draft emails, posts, and other content quickly with AI as a starting point.
- Edit AI output so it sounds human, on brand, and like the client.
- Capture a client's voice so AI drafts sound like them, not a generic bot.
- Spot and fix the signs of generic AI voice before anything goes out.
Lesson 2.1 Drafting Fast Without Losing Quality
AI should give you a head start, never a finished product. Use it to get past the blank page, then bring your own editing eye.
- Start with AI for the first draft, not the final version.
- Give it examples. Paste in three to five real samples of the client's past writing so it has a voice to match.
- Ask for options. Request three to five variations at once, then pick and combine the best lines.
- Set a hard rule: no AI draft goes out unedited by you.
Lesson 2.2 Capturing a Client's Voice
Generic AI output happens when AI has nothing real to imitate. Give it the client's actual voice and the output improves immediately.
- Collect examples: save five to ten of the client's best past emails, posts, or captions.
- Note their habits: short sentences or long, emoji or none, formal or casual, favorite phrases.
- Paste examples into every prompt: include two or three real samples so AI has something concrete to match.
- Sanity check the voice: ask AI to describe the voice back to you in a few words, then correct anything it gets wrong.
Lesson 2.3 Killing Generic AI Voice
Clients notice generic AI writing fast, and it damages trust in your work. Learn the tells and edit them out every time.
- Watch for tells: phrases like "in today's fast paced world" or "unlock your potential," and excessive exclamation points.
- Cut the throat clearing. Drop the generic opener and start with the actual point.
- Add a specific detail only the client would know, a real product name, a real number, a real event.
- Read it out loud. If it does not sound like a person talking, rewrite the sentence.
Five captions, edited into brand voice
- You ask AI for five caption options for a new product photo, giving it two past captions as voice examples.
- Three of the five sound generic and overly excited, so you cut them.
- You keep the two strongest, then trim the exclamation points and add the client's usual sign off.
- You add one specific detail, the actual scent name, that AI could not have known on its own.
- You send the client two polished, on brand options in under ten minutes.
Build your content workflow
Content prompts
Ready-to-use prompts for the writing tasks you will do most.
A client tells you the AI-written posts you have been sending sound generic. What is your best fix?
Module 2 Quiz
What should an AI draft always be treated as?
What is the best way to help AI match a client's voice?
Which of these is a sign of generic AI voice?
A client says AI-written posts sound generic. What is the right response?
Module 3: Research and Summarizing
- Use AI to research faster while still checking your sources.
- Summarize long documents and call notes into something a busy client will actually read.
- Extract data from messy text into a clean, usable format.
- Build a habit of verifying anything AI tells you before you rely on it.
Lesson 3.1 Researching With AI and Vetting What It Gives You
AI can get you oriented on almost any topic in minutes. The skill is knowing when to stop trusting it and start checking.
- Get oriented fast: use AI to learn key terms, common approaches, and the right questions to ask next.
- Separate claims from sources: ask AI to list its claims separately from where they came from, so you can check each one.
- Treat it as a starting point, then confirm key facts on the actual source.
- Never repeat a statistic, price, or claim to a client until you have seen it somewhere real.
Lesson 3.2 Summarizing Long Documents and Calls
Busy clients do not want the whole document, they want what it means for them. AI can turn hours of reading into a page they will actually read.
- Ask for a one page brief: three key takeaways, followed by a short list of action items.
- Feed it the real content. Paste in the actual document or transcript rather than asking AI to guess from a topic.
- Ask it to flag gaps. Have it note anything unclear or missing so you know what to follow up on.
- For call notes, ask for decisions made, action items with owners, and open questions.
Lesson 3.3 Extracting Data Into a Usable Format
Messy information becomes useful the moment it lands in a clean table. AI is fast at this, but it still needs a check before you trust it.
- Paste in the mess: an email thread, a PDF, or a spreadsheet dump, and ask for a clean table.
- Specify exact columns, for example name, date, amount, and status.
- Spot check a few rows against the original before trusting the whole extraction.
- Save it straight into a spreadsheet or the client's tracker.
A 20 page report into a one page brief
- The client sends a 20 page vendor report and asks "what do I need to know."
- You paste the report into AI and ask for three key takeaways and a list of action items.
- AI's summary mentions a savings figure, so you open the report and confirm the exact number.
- You format the brief as one page: takeaways at the top, action items with owners below.
- You send it to the client with a note on the one thing that needs a decision this week.
Practice research and summarizing
Research and summary prompts
Prompts for research, summaries, and pulling clean data from messy text.
While researching for a client, AI cites a specific statistic with no source given. What do you do?
Module 3 Quiz
What should you do with a statistic AI gives you before sharing it with a client?
What makes an AI summary most useful to a busy client?
When extracting data from messy text into a table, what should you always do first?
What is the biggest risk of treating AI research as finished work?
Module 4: Automation and Workflows with AI
- Spot which repetitive tasks are actually worth automating.
- Use the AI features already built into tools you use every day.
- Build a simple AI-assisted workflow from trigger to output.
- Measure the time an automation actually saves so you can prove its value.
Lesson 4.1 What Is Worth Automating
Not every repetitive task should be automated. The best candidates are predictable, low risk, and happen often enough to matter.
- Look for tasks you do the same way, more than a few times a week.
- Look for a clear input and output, not tasks that need constant judgment calls.
- Skip anything high stakes or public facing unless it has a human check step built in.
- Start with one workflow, prove it works, then expand.
Lesson 4.2 AI Features Already Inside Your Tools
You do not need to build anything complicated to start. Most tools you already use have AI features sitting one click away.
- Gmail and Outlook: AI drafted replies and summarized email threads.
- Notion and Google Docs: built-in summaries, outline generation, and rewriting tools.
- Zapier and Make: connect an AI step between two apps without writing code.
- Calendar tools: AI scheduling assistants that suggest meeting times.
Lesson 4.3 Building a Simple AI-Assisted Workflow
Every good AI workflow has the same four pieces. Define them clearly and the workflow stays reliable.
- Trigger: what starts the workflow, a new form entry, a meeting ending, an email arriving.
- AI step: what you ask AI to do with that input.
- Human check: where you review before anything goes out.
- Output: where the result lands, an email sent, a task created, a document updated.
Meeting notes into a task list and a client email
- Trigger: a client call ends and you have a raw transcript.
- AI step: you ask AI to pull out decisions made, action items with owners, and open questions.
- Human check: you read it against your own notes and fix two mislabeled action items.
- Output: you turn the cleaned list into a task list in your project tool and a short recap email to the client.
- Total time: about ten minutes, down from thirty spent writing notes by hand.
Design your first AI workflow
Workflow plan template
Map any AI-assisted workflow in four fields before you build it.
An AI-assisted workflow you built has run smoothly, but it occasionally produces a wrong output. What is your best fix?
Module 4 Quiz
What kind of task is the best candidate for automation?
Which of these is an AI feature already built into a common tool?
In a simple AI-assisted workflow, what comes right before the output?
An AI automation occasionally produces a wrong output. What is the right fix?
Module 5: AI for Admin and Customer Service
- Draft support replies quickly without losing a human tone.
- Use AI to triage and prioritize a busy inbox.
- Keep a warm, human tone in every AI-assisted reply.
- Protect client and customer data when using AI for support work.
Lesson 5.1 Drafting Replies and Macros Fast
Support work rewards speed, but never at the cost of sounding like a robot. AI can draft the reply, you make it real.
- Build a library of your five most common reply types, then ask AI to draft variations from each.
- Give AI the actual message so the reply answers what the customer actually asked.
- Always personalize the name, order details, and specific issue before sending.
- Keep a short list of things AI should never say without your review, like refunds or policy exceptions.
Lesson 5.2 Triaging an Inbox With AI
A full inbox is overwhelming until it is sorted. AI can scan and rank messages in seconds, so you spend your time on what matters most.
- Sort by category: urgent, routine, and can wait.
- Flag sensitive language: anything mentioning a refund, a complaint, or legal terms goes to your eyes first.
- Summarize the queue: ask for a one line summary of each open ticket so you can scan fast.
- Set the rule: AI sorts and drafts, you decide and send.
Lesson 5.3 Scheduling and Note-Taking Help
AI is especially good at the small admin tasks that eat up your day in five minute pieces.
- Turn a messy back and forth into a clear meeting time proposal.
- Summarize a call into notes and next steps right after it ends.
- Draft a calendar invite description from a few bullet points.
- Double check any date, time, or time zone AI proposes before you send it.
Three support replies, personalized in minutes
- Three tickets come in: a shipping delay, a sizing question, and a billing complaint.
- You paste each customer message into AI with your reply template and ask for a draft.
- You personalize each draft with the customer's name and specific order details.
- You cut one line from the billing reply that promised something outside your authority.
- All three go out within fifteen minutes, each sounding like a person, not a script.
Build your AI-assisted support kit
Support prompts
Prompts for support replies, inbox triage, and meeting notes.
AI drafts a support reply that shares more detail than the client would want made public. What do you do?
Module 5 Quiz
Before sending an AI drafted support reply, you should always?
What is the best use of AI for inbox triage?
What is the rule for keeping AI-assisted replies human?
An AI drafted reply shares more information than the client would want public. What should you do?
Module 6: Using AI Responsibly and Selling It
- Build a verification habit you use on every AI-assisted task.
- Set a clear rule for protecting client and customer data around AI tools.
- Explain the real value of AI-enabled VA work in plain terms.
- Charge a premium for the speed and output AI lets you deliver.
Lesson 6.1 Accuracy and Verification Habits
Every AI-enabled VA needs one non negotiable habit: check before you send. This single habit protects your reputation more than any tool choice.
- Treat every AI output as a draft from a fast but occasionally wrong assistant.
- Verify facts, numbers, and sources before anything reaches a client or their customer.
- Read every AI draft out loud once before sending it.
- Keep a personal checklist of mistakes you have caught before, and check for them each time.
Lesson 6.2 Privacy and Client Data
Clients trust you with sensitive information. That trust does not automatically extend to every AI tool you might open.
- Never paste private data such as financials, passwords, or customer lists into a free public AI tool.
- Check the tool's settings. See whether it trains on your inputs, and turn that setting off when you can.
- Ask each client what is and is not okay to put into an AI tool before you start.
- Keep a simple written rule you can show a client if they ask how you protect their data.
Lesson 6.3 Positioning and Pricing Your AI Skills
AI is a tool, your judgment is the product. Sell the result AI lets you deliver, and price for it accordingly.
- Do not sell "I use AI," sell the outcome: faster turnaround, more output, the same quality.
- Name a specific result, for example "what used to take two days now takes half a day."
- Charge for the value delivered, not the tool used, your judgment and review are the paid skill.
- Offer an AI-enabled tier of your services at a higher rate than your standard tier.
Your AI-enabled tier is not cheaper for the client because it was faster for you. It is priced for the outcome they receive.
Turning "I use AI" into a client offer
- Weak pitch: "I use AI to help with my work." Vague, and it does not say what changes for the client.
- You rewrite it around the outcome: turnaround time, output volume, and your review process.
- New pitch: "I use AI to draft your content and support replies fast, then I review and polish everything myself, so you get a full week of content in about half the time."
- You offer this as a premium tier priced above your standard hourly rate.
- The client books the higher tier because the pitch is about their result, not your tool.
Build your responsible AI practice
AI-enabled service offer template
Turn your AI skills into a priced offer a client can say yes to.
A prospective client says they are worried using AI means lower quality work. How do you respond?
Module 6 Quiz
What should you do before sending any AI generated fact or number to a client?
What is the safest rule for client data and AI tools?
What should you actually sell when pitching your AI-enabled services?
A client worries that using AI means lower quality work. What is the best response?