How to Make Presentation Slides with AI: Tools, Workflow, and Prompts

Your presentation is first thing tomorrow morning — and your slides are still blank. On a night like that, you can now type a single line of theme and, minutes later, have 20 draft slides lined up in front of you. In 2026 that isn't an exaggeration. AI has taken over the heavy parts of building a deck — thinking through the structure, polishing the design, and writing the script — all at once.

Here's the bottom line. If you want to reach a "show-able draft" as fast as possible, throwing your theme at an all-in-one generator like Gamma is quickest. If your final deliverable must be PowerPoint (.pptx) and you absolutely want to avoid broken layouts, the Copilot built into PowerPoint is the solid choice. And the single thing I most want to convey in this article: treat "writing the words" and "designing the slides" as separate jobs, and AI slides suddenly come together. Structure goes to an LLM (like ChatGPT); design goes to a dedicated tool. This division of labor is the winning move of 2026. Below: what AI can automate, the two approaches, a tool comparison, a hands-on workflow, prompts, tips, and pitfalls.

* The tool specs, pricing, and features in this article are based on vendor-published values and several comparison outlets (as of 2026). AI moves fast, so check each official source for the latest. AI-generated figures, quotes, and sources can be wrong — always verify anything important yourself.

1. What part of slide-making can AI automate?

"AI slides" actually spans three stages. Which one you want to hand off changes which tool you pick.

• ① Structure (outline): what to say, in what order, in how many slides. The skeleton of the talk.
• ② Script / body: the wording on each slide, the bullets, and the speaker notes (what you'll actually say).
• ③ Design: layout, color, fonts, diagrams/charts, images. The visual finish.

It used to be that people did ① through ③ entirely by hand — which is why a single deck could swallow hours. In 2026, AI can automate nearly all three stages, to a "draft" level. But — and this is the key point — it's a "draft," not a "finished product." As we'll see, AI's first cut has too much text and sometimes shaky facts. The job of finishing is still a human one. AI builds 0→70 in minutes; the person polishes 70→100. That, I believe, is the correct split.

2. Two approaches: all-in-one vs. division of labor

So how do you run those three stages? There are two broad routes. Think of them as differing in where you draw the line between "draft" and "finish" from the previous section.

Personally, I use them by case: A (all-in-one) is plenty for a light internal share or a study session, while B (division of labor) is for moments that hinge on substance — a client pitch, a board report. And the two aren't mutually exclusive. The strongest combo is actually to lock down the script in B, then pour it into an A-type tool to design. The workflow in the next section is exactly that.

3. Comparing the major tools

Here are the representative AI slide tools, lined up by strength. There's no "all-in-one champion"; the 2026 knack is choosing by the exit — are you delivering in PowerPoint, or sharing on the web?

* Features, pricing, and support are published values as of 2026. Vendors update frequently. Pricing changes easily, so check the official source before adopting.

Here's one realistic lesson. Tome — once called "the face of AI presentations," with 20 million users — shut down its slides feature in April 2025. A trendy tool isn't necessarily the standard — which is exactly why the division-of-labor mindset of "keep the script (as text) in your own hands" pays off, rather than over-depending on any one tool. Even if the tool disappears, the content survives.

4. A practical workflow: a finished deck in 5 steps

Building on the comparison, here's the most repeatable "division-of-labor × all-in-one combo," in five steps. Once you're used to it, the draft takes under ten minutes.

The point is to finish the "words" in STEPs 1→2, and only then hand them to design in STEP 3. Most people dump everything onto an all-in-one tool and then burn time fixing the half-baked content it returns. Lock down the content first, and the design tool's only job becomes "arrange it nicely." That's the practice of "split writing from design."

5. Copy-paste prompts

Here are prompts that work for STEPs 1 and 2. The more you hand over "audience, goal, time, tone," the closer the output gets to usable.

① Build the outline (structure)

# Create a presentation structure
You are an experienced presentation designer. Build a slide structure under these conditions.

- Theme: [e.g. internal proposal for a new service]
- Audience: [e.g. 5 executives, not AI-savvy]
- Goal: [e.g. get budget approval]
- Time: [e.g. 10 minutes, about 10 slides]
- Tone: [e.g. concise, logical]

Output: per slide, a table of "title / the one thing to convey / key points (up to 3)."
At the end, add 3 likely objections from the audience, with answers.

② Flesh out one slide + speaker notes

# Flesh out a slide
Finish "Slide 3" above.
- Body: up to 3 bullets, each under ~10 words (to be shown, not read)
- If you include a number, mark it [needs source] so I can verify later
- Separately, write "speaker notes (60 seconds)" in spoken style, ready to read aloud

③ Reformat for a design tool

# Convert into a format to paste into Gamma etc.
Reformat this structure so it's easy to paste into an AI slide tool.
- Separate each slide as Markdown: "heading -> bullets"
- One message per slide. Minimal text
- Where something could be a diagram, note "diagram here (comparison/process/figures)"

By flagging "where something could be a diagram" as in ③, the design tool finds it easier to turn prose into visuals. Showing with diagrams and keywords — not a wall of bullets — is the first step to slides that land.

6. Six tips for slides that land

The know-how to lift AI's draft up to a presentable level. In order of impact.

Of these, the one that transforms beginners most is ② "cut the text in half." AI's draft is, almost without exception, text-heavy. Cut boldly, and move the deleted sentences into your speaker notes. That alone turns "slides you just read aloud" into "slides that make the audience look up."

7. Pitfalls (broken layouts, bloat, fabricated data)

For all the convenience, AI slides have classic pitfalls. Know them and you'll avoid 90%.

• Layouts break on .pptx export: web-based generators (like Gamma) tend to shift margins and fonts when exported to PowerPoint. If your final deliverable must be PowerPoint, choose Copilot (built into PowerPoint) or a native-.pptx tool from the start — that's the safe move.
• The first draft is wordy and bloated: AI's body text tends toward too many bullets. Budget 30–60 minutes to "cut and visualize." Generation saves time; revision never drops to zero.
• Plausible fabrication (hallucination): AI sometimes naturally invents stats, quotes, and sources. When a specific figure or "per X Corp's survey" appears, always check the original. Getting called out on the day destroys trust in an instant.
• Handling confidential information: pasting unreleased figures or customer data into a cloud AI means sending it externally. Check your company's rules and each service's data handling policy. If concerned, have it build only the structure with names and figures redacted.
• Tool dependence: as with Tome's shutdown above, tools can disappear. Keep your script as text in your own hands, and build in a way that isn't bound to one specific tool.

Honestly, the two biggest risks of AI slides are "fabricated data" and "the laziness of shipping it without cutting." Put the other way: keep just two habits — "check numbers against the original" and "cut the text in half" — and AI slides become a weapon you can trust.

8. Recommendations by use case

When in doubt — first try Gamma's free tier to feel "how fast AI slides are," then switch to Copilot or division of labor once you learn PowerPoint delivery is required. That order wastes the least time.

Summary

AI slide-making reached, in 2026, a level that "builds a draft in minutes." Here's the gist.

• What AI can take on is the "draft" of structure, script, and design. The finish (cutting, verifying) is the human's job.
• The strongest is dividing "writing" from "design." Content by the LLM, looks by a dedicated tool. Keep the script as text in your own hands.
• Choose by the exit. Web sharing → Gamma; PowerPoint required → Copilot; Google co-editing → Gemini; beauty → Beautiful.ai; assets → Canva.
• The more you give "audience, goal, time, tone" in the prompt, the higher the quality of the first draft.
• Two habits prevent accidents: check numbers against the original / cut the text in half.

In the end, AI isn't "the person who makes the slides" — it's "the partner that produces a draft in an instant." The time spent frozen in front of a blank slide is over. But the final push that moves an audience — which number to emphasize, what to cut, how to tell it — is, now and in the future, a job only you can do.

FAQ