You will be able to use AI features in presentation software and chart tools without breaking brand rules or accuracy.
Open your presentation app, find its AI feature, and check your company policy to see whether that feature is approved for work data. The common ones are Copilot in PowerPoint and Gemini in Google Slides. You will only see them if your company has switched them on. A tool being built into an app you already use does not by itself mean the company has approved it, as Module 2 explains for any tool. If the policy is unclear, ask before you paste in anything from work.
Presentation apps increasingly have AI features that draft slides from an outline or a document, suggest layouts and rewrite text to fit the space. There are also separate AI presentation tools that build a whole deck from a prompt. These tools can save real time. For a rough internal deck, they are a quick way to get from a storyline to something presentable. For anything going to management or clients, generated decks tend to need more fixing than they first appear to. A deck can look finished and still be wrong in ways that are easy to miss.
Kavitha, an invented example, had already finished her storyline and speaker notes. She pasted her outline into her presentation app's AI feature and pressed generate. In under a minute she had twelve slides with photographs, icons and a colour scheme. It looked impressive for about five minutes. Then she noticed that the fonts were not the company's and the logo was missing. One slide had a photo of a smiling office team who did not exist. A chart showed competitor prices on an axis starting at S$400, which made a small difference look enormous.
Most of the trouble with generated decks comes from two areas: the company template and charts. Images need care as well.
Generated layouts often ignore the company template. They pick their own fonts, colours and layouts, leave out the logo or the confidentiality footer, and arrange content in ways the brand guidelines do not allow. Bringing a generated deck back into line can take longer than building it properly in the first place.
Many teams work the other way round. They use the AI assistant to draft the words, as in Lessons 6.1 and 6.2, and paste the approved text into the company template. The template supplies the design and the assistant supplies the content, so nothing needs undoing. If your app's AI can work inside the company template, try that too, then check the result against the template slide by slide.
Do not let a design tool invent a chart from a description. Build every chart from actual data in one of two ways. You can make it in a spreadsheet, where the chart is linked to cells you can check, or you can ask an assistant for chart code and run that code on your own data file. Either way, the numbers come from the source and not from the model's reading of your notes.
Check every chart as carefully as you check figures. Four checks catch most problems:
Axes: a bar chart should start at zero, and the scale should be sensible. Labels: the categories, units and time periods should be correct. Totals: the bars or segments should add up to what the source says. Chart type: the chart should show the point made in the slide headline and not some other point.
A truncated axis is an axis that does not start at zero. It can make a small difference look huge. For example, a five per cent difference can look like a doubling. Kavitha rebuilt her pricing chart in her spreadsheet with the axis starting at zero. The difference between competitors was still visible, but now it looked its actual size. The slide was more honest, and it also turned out to be more persuasive, because the finance head did not have to question it. An honest chart with its axis at zero can be more persuasive in general, because the audience has no reason to question it.
AI tools can produce convincing images of offices, people, products and places. Generated images can be fine as obvious illustrations, where the company allows them. They become a problem when they are presented as real. Do not use an AI-generated image of a real person, your real product, a client's site or a real place as though it were a photograph. Suppose a generated picture is labelled "our Jurong warehouse" but is not the actual Jurong warehouse. That picture misleads the audience, and if anyone notices, it damages trust in everything else on the slide. The same rule applies to invented customers, made-up screenshots and products shown with features they do not have.
Go through your deck and remove any generated image presented as a real person, product, client site, place, customer or screenshot. Replace it with a real photo, a plain diagram or nothing at all. Kavitha deleted the generated photo of the smiling office team and put nothing in its place, and the slide was clearer without it.
Your next step is to take three slides from your own storyline into your company template. Make one of them a chart built from real data, and go through each figure against its source.
Build three slides from your storyline in your company template, including one chart made from real data, and check each figure against the source.
Junxiong-WFG Organisation is an authorised representative of AIA Financial Advisers Private Limited (Reg. No. 201715016G).