How to Use AI to Improve Your Written Communication as an Executive

Start by giving AI a piece of writing you are not satisfied with and asking for specific diagnostic feedback.

'Here is an email I wrote: [paste]. What are the main weaknesses in this communication? Be specific and direct.'

Common diagnoses AI provides: Buried lede: the most important point is not in the first sentence. The reader has to read the whole communication to understand what you want. Passive voice: 'It was decided that...' rather than 'We decided...' creates distance and obscures accountability. Jargon and nominalisations: converting verbs to nouns ('the realisation of benefits' rather than 'benefiting') makes writing harder to read without making it more authoritative. Unclear ask: the communication does not specify what action the reader is being asked to take.

Once you understand your specific patterns, you can target improvement. Ask: 'What is the single most common weakness in my writing based on these samples?' and focus on that one thing first.

For immediate improvement on a specific piece of writing:

Conciseness: 'Reduce this to half its length without losing any of the essential content.' This forces the AI to identify which words are adding nothing and which are essential. If the reduced version is better, that tells you something about your original.

Clarity: 'Rewrite the first paragraph so that the key message and the action required are in the first two sentences.' This is the 'pyramid principle' applied: lead with the conclusion, support it subsequently.

Audience appropriateness: 'This was written for [audience]. Rewrite it for [different audience], adjusting the level of detail, the assumed knowledge, and the tone.' Being explicit about audience forces better calibration.

Persuasiveness: 'Rewrite this proposal to be more persuasive for a board that is sceptical of [type of investment]. Lead with the business case, not the process.'

The most effective use of AI for writing improvement is building it into a personal editing process, not using it as a last-minute fix.

A practical editing process for important communications: Step 1: write your first draft without AI assistance. The first draft is your thinking; outsourcing it to AI from the start produces generic communication. Step 2: use AI to diagnose weaknesses: 'What is unclear? What is the main weakness? What would a sceptical reader miss or misunderstand?' Step 3: edit yourself, using the AI feedback as a guide but making your own writing decisions. Applying AI suggestions without understanding them does not improve your writing; it just improves this piece. Step 4: for important external communications, use AI to check tone and completeness before sending: 'Is the tone right for [audience/situation]? Is there anything missing that this recipient would expect to see?'

This process improves both the specific piece and your writing capability over time. The diagnosis stage builds your awareness of your patterns; the self-editing stage builds your skill.

AI can accelerate writing development beyond just fixing individual pieces.

Ask for the principle behind a specific suggestion: 'You suggested I change [X] to [Y]. Why is [Y] better? What writing principle does this illustrate?' Understanding the principle, not just adopting the suggestion, builds capability that transfers to future writing.

Analyse writing you admire: 'Here is a piece of writing I find particularly effective: [paste example]. What techniques make it effective? What specific features could I adopt in my own writing?'

Request structured writing coaching: 'Based on the writing samples I have provided, what are the three specific habits that would most improve my executive communication? For each, give me a concrete practice or exercise.'

The leaders who improve most from AI writing assistance are those who use it as a learning tool, not just an editing tool. The goal is writing that reflects your thinking clearly, not AI writing that happens to come out under your name.