Many thought AI would be a magic fix that would save time and make us sound more ‘professional’. But today’s readers can spot weak AI business writing in seconds.
So every time you treat AI’s first draft as the final draft, you’re putting your professional reputation on the line.
On 3 June 2026, we ran a live webinar on AI writing risks and remedies – drawing over 450 registrations. That tells us that many professionals are wrestling with the same thing: How do you use AI to write faster without losing your voice and credibility?
This article walks you through the seven credibility killers we covered in our webinar – and how to fix them. Watch the full recording below. Or read on.
First, two cautionary tales
In October 2025, Deloitte delivered an AI-assisted report to the Australian government that contained made-up citations, dodgy footnotes and fabricated quotes. It made headlines around the world.
But blaming AI misses the point. The real failures were the people who trusted the output blindly – and leaders who never put AI writing guidelines in place to catch it.
Most of us won’t make headlines for an AI blunder. But we can still damage our professional reputation every time we publish AI writing that doesn’t sound like us.
In another case, thought leader Tony Robbins shared a LinkedIn post that seemed to come straight out of a chatbot. Readers flagged and mocked his post as AI. And his credibility was instantly compromised.
We have a name for that type of writing in our workshops: ‘word salad’. Lots of words on the page that say nothing of value.
CREDIBILITY KILLER 1: Repetitive phrasing and structures
AI learned to write by consuming millions of documents – and it absorbed every cliché along the way. So unless you tell it otherwise, it defaults to those predictable patterns.
Here are a few telltale tropes you might recognise:
- The forced negative: It’s not just about X, it’s about Y…
- Rhetorical questions before every new point: The real impact?
- Empty phrases: To drive meaningful outcomes
- Patronising drama: Let that sink in
- LinkedIn-guru transitions: Here’s the thing…
Why it kills credibility: When your writing reads from the same template everyone else is using, people assume you outsourced it. They start questioning your authenticity AND your competence.
How to fix it: Edit hard and prompt against the patterns. Give your tool a list of banned phrases. Then keep reminding it – because it will spring back into old habits. Better still, keep a list of AI words and phrases with your team and update it as new ones appear.
CREDIBILITY KILLER 2: Writing for no one
AI doesn’t know your audience like you do.
AI only knows what you tell it. If you don’t provide the context around your audience – what they already know, what matters to them and what you want them to think, feel or do – it fills in the gaps with generic language. And generic writing rarely connects.
Why it kills credibility: Readers want you to acknowledge their world and the relationship you share – whether that’s a colleague of one week or 10 years. People who don’t feel seen place their attention and trust elsewhere.
How to fix it: Get clear on your audience before you open your AI tool. Once you know who you are writing to, put all that context in your prompt: their role, what they know, what they care about, what frustrates them and the nature of your relationship.
CREDIBILITY KILLER 3: A stiff, wordy tone
AI was trained heavily on polished corporate and academic writing – and that type of writing is overly formal by nature. So when you prompt AI to ‘make it professional,’ it hears ‘make it formal.’
But the best business writing is concise and conversational.
Why it kills credibility: Stiff, formal writing builds a wall between you and your reader – when you actually want to build is connection. How do you connect with readers? Through plain, conversational language that’s easy to follow.
How to fix it: Cut ruthlessly and write more like how you talk. Watch your sentence length, too – comprehension drops dramatically when sentences are over 25 words. Learn the power of plain English so you can apply the principles.
CREDIBILITY KILLER 4: Burying the point
AI cannot convey your key message unless you’re clear on it yourself. Left to fill the gaps, AI makes assumptions, draws its own conclusions and emphasises the wrong parts.
Worse still, you might not even notice these gaps if you skipped the underlying thinking that would have caught them.
Why it kills credibility: Readers trust people who respect their time. If they have to go gold mining for your point, they get frustrated, switch off and think twice before opening your next message.
How to fix it: Lean on the inverted pyramid of writing: lead with the must-know, follow with the helpful-to-know and cut the nice-to-know. Check whether someone could grasp your key message in 10 seconds or less.
CREDIBILITY KILLER 5: The passive problem
Passive sentences either leave out who’s doing the action (‘The report was written’) or put them after the action (‘The report was written by Jo’).
The active version is simply ‘Jo wrote the report.’
Passive voice is everywhere in the academic papers and corporate documents AI learned from. So unless your prompt makes clear who did what, AI often falls back on passive constructions too.
Why it kills credibility: Passive writing is weaker, wordier and more convoluted. Plus, it strips your message of ownership and responsibility.
How to fix it: Learn to write actively to make actions clear. And know the few places where passive voice still earns its keep, such as legal contexts.
CREDIBILITY KILLER 6: Hedging
Hedging is the language of doubt: may, might, probably, it seems and to some extent.
These words and phrases let you say something without committing to it. AI loves hedging because it’s trained to sound safe and balanced. It’s trained to avoid being wrong.
Why it kills credibility: Excessive hedging makes readers question your expertise. If you’re not confident in what you’re saying, why would they be confident in you?
How to fix it: Be decisive. If you have evidence for your recommendation, state it plainly. Delete qualifiers unless there is a genuine reason to keep one, such as HR matters.
CREDIBILITY KILLER 7: The wall of text
Readers judge your writing before they read a single word.
A dense email or document full of wall-to-wall text feels like hard work – making it far more likely to be postponed or ignored.
Yet this is one of AI’s biggest blind spots. It optimises for thoroughness over readability, so it often produces exactly the kind of dense writing readers avoid.
Why it kills credibility: A wall of text signals disorganised thinking. There’s a compound risk too: When a document looks like hard work, people drop it into AI and ask for a summary. If your thinking was muddy going in, something even muddier comes out.
How to fix it: Learn to love white space to make skimming easy. Use sub-headings, short paragraphs and plenty of bulleted lists. Bold the date or action so scanning eyes can find it.
What to do next
If you lead a team:
- Create and embed a writing style guide (if you haven’t already). This document sets clear expectations for how your organisation communicates, so your team’s writing is consistent regardless of who’s behind the keyboard.
- Give your team the business writing training they need. Poor AI writing almost always comes from people who never learned what good writing looks like – so they cannot spot what to fix.
- Introduce AI writing guidelines. These are standard operating playbooks covering which tools to use, where AI fits in your process and what never goes into it. Your people are already writing with AI – so give them the rules to do so safely.
If you write for yourself:
- Match the tool to the task. A free grammar checker is fine for polishing a routine email. But if you’re drafting a proposal, report or other high-stakes document, choose a tool that challenges your thinking rather than simply generating text.
- Configure your AI. Give it permanent context about who you are, who you write for and how you like to communicate. It will save you time and produce more useful first drafts.
- Strengthen your editing skills. The first draft may be AI’s job. But the final draft is yours. Learn to spot weak structure, generic language and the subtle changes that undermine your message.
The bottom line
AI is a useful first-draft assistant. But it’s a terrible final author.
Human judgement and a sharp editing eye turn generic drafts into writing you’re proud to put your name to.
Those skills are learnable – and exactly what we teach. Explore our Mastering AI Writing course for teams and Mastering AI Writing course for individuals today to learn more.
Your writing leaves fingerprints. Make sure they are yours.