'You Only Need to Explain Why You DIDN'T Use AI'

‘You only need to explain to me why you DIDN’T use AI.’

Not my words but those of the CEO of the company that owns major media outlets like Politico.

This bullish position strikes at the heart of a major modern conundrum. Should we use AI and if we do - should we tell people?

The shape and scale of AI's impacts on our economy, information environment, and climate aren't yet clear, but they will be significant. Efforts by governments to plan for AI seem to take a narrow view of a wide range of possible consequences.

Consider what's already happening.

AI-generated content floods social media without disclosure, making it impossible to distinguish authentic voices from synthetic ones. Businesses quietly use AI for customer service, research, and content creation while data and profits continues to be hoovered up by a small number of the most powerful companies in history.

The jury remains out on AI's practical benefits - one recent study claimed 95% of AI pilot projects end in failure.

But that uncertainty makes transparency even more crucial.

I've been running my own pilot over recent weeks: creating a newsletter that uses AI daily to analyse news and offer counter-narratives to populist messaging. (You can ask for the link in the comments!)

People want to make informed choices about AI-mediated content.

AI now builds whole websites in minutes. But you wouldn’t always know.

Transparency isn’t always straightforward.

How do you label research where AI helped with initial data gathering but humans did all the analysis? Or presentations where AI suggested structure but humans crafted every argument?

Or vice versa?

Are catch-all disclaimers in website and email footers ok? Or do we need to be more specific?

Flagging the use of AI and being willing to share and learn from each other is important.

Especially when money is tight for many organisations.

My newsletter uses AI every day to analyse the news and offer thoughts on how to respond.

I write my own Linked In posts but often get a second opinion on them from AI.

Should we assume AI is always being used?

Or always acknowledge it?

Take the poll.