A speech that resonates

How can you use AI to build a clear, credible speech that feels natural to deliver?

1250 € per session, excl. VAT

In short

This half-day session of 3 hours turns an intention to speak into a speech that can be delivered aloud. Using an approved generative AI tool, participants define their message, organise their arguments, draft and then rehearse a short speech. They leave with an annotated text and a completed quality assessment grid.

Why this training?

A speech written like a document often proves difficult to listen to: long sentences, arguments piled on top of one another and an imprecise conclusion. Preparing to speak requires selecting a guiding idea and making its development clear on the first hearing. AI can suggest outlines and wording, but it can also introduce invented facts or an artificial tone. Participants remain responsible for the evidence, the accuracy of the message and the words they agree to say aloud. Continuing case study: The fictitious association « Passerelle des Savoirs » is opening a community repair workshop. Its coordinator must deliver a four-minute speech to volunteers and local residents: explain the project, reassure beginners and invite everyone to an introductory session. A fictitious briefing pack provides the only facts that may be used. Application at work: For their next speech, participants reuse the planning sheet, share only approved information with the AI and rehearse with a timer. They then ask a listener to restate the message they remember and the action expected, before adjusting their text.

Objectives of this training

  • Formulate a central message and an expected action for a defined audience
  • Compare two outlines suggested by AI against the purpose of the speech
  • Write a four-minute speech based on supplied facts
  • Revise the text to make it easier to understand when spoken
  • Evaluate the speech through timing and feedback from a listener

Who is it for and what are the prerequisites?

Roles and tasks concerned

  • Managers who need to open a meeting or present an initiative
  • Association leaders and communications officers preparing a short public speech

Prior skills

  • Have previously entered a prompt and requested a revision in a generative AI assistant
  • Be able to describe the audience and purpose of a speech

What you will work on

  1. Define what the audience should remember

    Objective: Establish a purpose before requesting a text. Audience expectations, context of the speech, central message, desired action and available facts; limits on the data that may be shared.

  2. Choose a convincing line of argument

    Objective: Organise ideas to be understood in a single hearing. A concrete opening, a progression through no more than three ideas, verifiable examples, transitions and a conclusion that inspires action.

  3. Write to be heard

    Objective: Produce a text that stays true to the speaker’s voice. A writing prompt with context, short sentences, concrete vocabulary, rhythm; checks for invented content and excessive promises.

  4. Rehearse, measure and tighten

    Objective: Revise the speech based on its actual effect. Timing, intelligibility, ease of delivery and recall of the message; distinguishing between a stylistic opinion and an observed difficulty.

What you take away

  • Planning sheet: audience, purpose, message and constraints
  • Reasoned outline with an opening, transitions and a conclusion
  • Four-minute speech annotated for spoken delivery
  • Completed quality assessment grid and priority corrections

Method and assessment

  • Short teaching inputs followed by immediate practice using the continuing case study.
  • Individual work followed by work in pairs, with participants reviewing one another’s texts.
  • Guided group review: what works, what needs correcting and what remains a human decision.
  • Preparation: Arrange access to a generative AI assistant approved by the organisation Choose a fictitious or anonymised situation, without personal data or confidential information

Observable criteria: The central message fits into one sentence and the listener can restate it; Every factual claim is linked to the supplied briefing pack; The opening sets out what is at stake and the conclusion proposes an explicit action; The timed reading lasts between three minutes thirty seconds and four minutes thirty seconds; The final version incorporates at least two corrections arising from rehearsal.

Frequently asked questions

Do you need to learn the text by heart?

No. The annotated text helps with rehearsal; a keyword sheet can then support more natural delivery.

Can AI invent a personal anecdote?

It must not present an invention as a lived experience. Use a fictional example that is explicitly identified as such.

Will we work on body language?

The priority remains the content and its spoken delivery; body language is not practised in depth.

Can you use an existing professional speech?

Yes, after anonymisation and in accordance with internal rules; otherwise, the fictitious briefing pack allows you to complete all the exercises.

Last updated: