We’ve all been there.
Someone hands you content they’ve “written.”
You check it out. Impressive! Very detailed, smart, well-sourced; turned around quickly, too.
Hey, wait a sec’ – was that too quick? Hmmm – and what’s this with the weird formatting and slick sounding passages?
You know what happens next. A lot of second guessing. How much AI was used, how much of the content was human- generated?
It’s what I call the AI moment of truth. Former SAS CEO Jan Carlzon referred to these key moments when a company makes an indelible brand impression – good or bad – with a customer. Similarly, in PR we are creating these moments of truth when we overrely on AI in a careless way.
What can be done, how can you harness AI without tarnishing your brand? Should you even try?
The AI “Sniff Test”
AI surrounds us today and offers help at every turn.
This begs the question: is the content you are viewing, the advice you are getting, the result of AI or a real person?
To clear the confusion, people try to sniff out telltale signs, via AI truth detectors, especially for media facing pitches and articles, as we know AI-generated content will not fly with many editors.
But in the real world, it is not a binary choice. Some may use AI for fine tuning. Give me a better headline! Rewrite my conclusion, or this passage that is too technical.
So, how much AI help is allowable? Is it OK to use AI to write the whole flipping article – as long you fact check, add a few touches, and hide the AI tracks? But, seriously, when does covering the tracks become more work than writing the article?
And what if AI actually came up with something better than you? Would you be remiss in not presemting the better content?
I’ve seen some interesting exchanges recently about AI checkers flagging an article with over 61% gpt score. Should we fix the article, get it down to a smaller gpt score and call it a day?
Clearly, there are all kinds of questions for PR teams that want to harness AI in a responsible and smart way. There are not a lot of easy answers – and we know the calculus will change soon, with the rapid evolution of our AI ecosystem.
Meanwhile, I am thinking that it mighty not be bad idea to tag your content with a nutritional facts statement, like the image here identifying the content of your content – what is the “nutritional” valiue, is much of it ultraprocessed AI slop?







