Clarity Over Hype: Why How We Talk About AI Matters

By John

Spend five minutes in any AI conversation right now and you'll hear it: "revolutionary," "game-changing," "will transform everything." It's everywhere; in product launches, LinkedIn posts, boardroom pitches. And it's doing more damage than most people realise.

Hype isn't just annoying. It's a trust problem waiting to happen.

The cost of overclaiming

When AI is oversold as smarter, safer, or more autonomous than it is the gap between promise and reality eventually shows up. A chatbot that "understands" turns out to hallucinate. A hiring tool marketed as "bias-free" turns out to have blind spots nobody tested for. When that gap becomes visible, it doesn't just embarrass the person who made the claim. It erodes confidence in AI generally, for everyone, including the teams doing careful, responsible work.

This is why some of the most credible voices in AI have leaned hard in the opposite direction: precision over enthusiasm, caution over confidence, plain language over jargon. Not because AI isn't impressive, it often is, but because being specific about what a system can and can't do is the difference between a claim that holds up and one that collapses under scrutiny.

What this looks like in practice

Communicating about AI with clarity rather than hype isn't about being dull or overly cautious to the point of uselessness. It's a discipline:

Say what the system does, not what it might one day do

Name the limitations in the same breath as the capabilities

Avoid words like "intelligent" or "understands" unless you can defend them

Resist the urge to match a competitor's bold claim with an equally bold one of your own

This matters enormously for anyone deploying AI internally too. If a team oversells a tool to leadership, expectations get set that the tool can't meet. Then, when it underperforms, the instinct is often to blame the technology rather than the communication that set it up to fail.

Why this belongs in ethics training, not just marketing

It's tempting to file "tone of voice" under branding and leave it there. But how an organisation talks about its AI systems — internally and externally — is an ethical choice, not just a stylistic one. Overclaiming misleads users, customers, and regulators about risk. Underclaiming can hide real capability that needs proper scrutiny. Getting the language right is part of getting the governance right.

This is exactly the kind of judgement we help teams build: not just knowing the rules around AI, but developing the instinct to communicate about it honestly, even when a bolder claim would be easier to sell.

If your organisation is thinking about how it talks about AI to customers, regulators, or its own leadership this is a conversation worth having before the language gets away from you.

Further to this, do your own staff know what they can & can’t do with AI or are you asking them to use their own personal devices by not providing the correct training resource, AI platform and oversight support? Are you aware that “AI work slop” could exist in your business right now because of a strategy that evolves around not training or empowering staff with AI or giving them a route to report poor or biased outputs.

We can assist with that conversation as well as the required training. We can assist as you build your own core AI oversight and structure for your teams.