How to Write Emails Faster With AI

Step 1: Choose Your AI Email Tool

The first step is selecting an AI tool that fits your email platform and workflow. Each option has different strengths, and the right choice depends on where you spend your time.

If you use Gmail, Gemini is built in and handles basic drafting at no extra cost. Shortwave replaces the Gmail interface with an AI-native client that drafts, summarizes, and bundles threads automatically. Superhuman layers on top of Gmail with AI triage and keyboard-driven speed.

If you use Outlook, Microsoft Copilot is the most integrated option, pulling context from Teams, OneDrive, and Calendar when drafting replies. Fyxer is a lighter-weight alternative that focuses on drafting and inbox organization within the native Outlook interface.

If you want a platform-independent option, general-purpose AI tools like ChatGPT and Claude can draft emails when you paste the relevant context into a chat window. This approach works with any email provider but requires manual copy-paste for each message. For occasional use on important emails, this is perfectly effective. For daily volume, an integrated tool saves more time.

Step 2: Write a Clear Prompt

The quality of an AI-generated email depends directly on the quality of your prompt. Vague instructions produce generic output. Specific instructions produce messages that are ready to send with minimal editing.

A strong email prompt includes four elements: who you are writing to (their role and your relationship), what you need to communicate (the specific message, request, or information), what tone to use (formal, casual, friendly, direct), and what action you want from the recipient (schedule a meeting, review a document, approve a request).

For example, instead of "write a follow-up email," try: "Write a professional follow-up to a client named Maria who I met at a conference last week. Reference our conversation about migrating their analytics platform. Ask if she is available for a 30-minute call next week to discuss next steps. Keep it warm but concise." This level of detail gives the AI enough context to produce a draft that sounds intentional rather than templated.

Integrated tools like Copilot and Shortwave handle some of this context automatically by reading the existing thread, but even with integrated tools, adding specific instructions about tone and intent improves the output significantly.

Step 3: Provide Context From the Thread

Context is what separates a generic AI draft from a contextually accurate reply. When you are responding to an existing conversation, the AI needs to understand what has already been discussed, what questions were asked, and what commitments were made.

Integrated email tools handle this automatically. Shortwave, Superhuman, and Copilot read the full thread history before generating a draft, so their replies reference specific points from earlier messages. If someone asked three questions in their email, the AI draft will address all three rather than giving a vague response that misses the specifics.

When using a standalone AI tool like ChatGPT or Claude, paste the relevant portion of the email thread into the chat window before asking for a draft. You do not need to include the entire conversation history, just the most recent exchange and any earlier messages that contain important context. Summarize the broader situation in a sentence or two if the thread is very long.

The more context the AI has, the more accurate and natural the draft will be. Skipping this step is the most common reason people find AI email drafts unsatisfying, because the AI is essentially guessing at the context rather than working from it directly.

Step 4: Adjust the Tone and Length

Once you have a first draft, the next step is refining tone and length. Most AI email tools offer tone adjustment as a built-in feature. In Superhuman and Shortwave, you can click a button to make a draft more formal, concise, or friendly. In Copilot, the tone coaching feature suggests adjustments before you send. In ChatGPT or Claude, you can simply ask "make this shorter" or "make this more formal" as a follow-up prompt.

Length adjustment is equally important. AI tools tend to produce drafts that are slightly longer than necessary because they err on the side of completeness. In professional email, shorter is almost always better. Ask the AI to tighten the draft to three sentences if the message is routine, or five to seven sentences for more substantive correspondence. Removing unnecessary qualifiers and filler phrases from AI drafts is the single most impactful edit you can make.

Tone consistency matters when you send multiple emails per day with AI assistance. If your tone shifts dramatically between messages because you used different prompts, recipients may notice. Establish a consistent baseline by using the same tone instructions across similar types of emails, or choose a tool that learns your style automatically from your sent history.

Step 5: Review and Personalize Before Sending

Never send an AI-generated email without reading it first. This is the most important step in the entire process. AI drafts can contain factual errors, inappropriate assumptions, or phrasing that does not match your intent. A quick review catches these issues before they reach the recipient.

When reviewing, check for three things. First, accuracy: does the draft state anything that is factually incorrect or that misrepresents your position? Second, completeness: does it address every point the recipient raised and include all the information they need? Third, voice: does it sound like something you would actually write, or does it feel generically AI-generated?

Personalization makes AI-drafted emails feel human. Add a specific reference to something the recipient mentioned, include a detail that only you would know, or adjust a phrase to match your natural speaking style. Even small personal touches, like using the recipient's first name in a natural way or referencing a shared experience, transform an AI draft from functional to genuinely engaging.

This review step typically takes 15 to 30 seconds per email, which is a small investment compared to the several minutes saved by AI drafting. Over the course of a day with dozens of emails, the net time savings is substantial even with careful review.

Step 6: Build Templates for Recurring Emails

Many professionals send the same types of emails repeatedly: meeting confirmations, project status updates, introduction emails, follow-ups after no response, and thank-you notes after meetings. Rather than prompting the AI from scratch each time, save your best AI-generated patterns as reusable templates.

In integrated tools like Superhuman and Shortwave, you can save snippets and templates directly within the app. For standalone AI tools, maintain a document with your best prompts and the outputs they produced, organized by email type. When you need to send a similar message, start from the saved template and customize the details for the current situation.

Over time, your template library becomes a personal playbook that makes email drafting nearly instant for common scenarios. The AI still handles the heavy lifting of language and structure, but your saved templates ensure consistent quality and tone across every message type you send regularly.