Paris has a new literary scandal. For once, nobody slept with anybody. This autumn, Thélyson Orélien’s novel was one of the darlings of the rentrée littéraire, the season when French publishers release hundreds of new books and the prize juries pick their favourites. It made the first selection for the Prix Goncourt. In France, that’s the big one.
Then came the accusations. The novel, critics claimed, had been written with the help of artificial intelligence, and some passages were allegedly lifted from other writers. In late September, the Académie Goncourt removed it from its list.
I’ll confess something. When I first heard the story, I thought the book had been denounced anonymously. Very French. Very dramatic. So I asked AI to check my facts, and it told me I was wrong. The essayist Samuel Fitoussi has said publicly that he started the whole thing.
So the machine accused of writing a novel corrected the man writing about the machine. I’ll take the irony.
Whether Orélien used AI isn’t for me to judge. What interests me is the reaction, and the assumption behind it: if you used AI, you cheated.
Guilty as charged
English isn’t my first language. I left France in 1986 to learn it in six months. Forty years later, I’m still working on it.
Last year, I said it publicly on LinkedIn: I use AI. I wasn’t ashamed then, and I’m not now.
Here’s what I use it for. Translating between French and English. Tidying my grammar. Research, as you’ve just seen. Illustrations for my LinkedIn posts and my blog, because nobody needs to see my drawings. And SEO blog posts for my wife’s jewellery business, because Google wants volume and I have a day job.
What I don’t let it do is write this. Or my LinkedIn posts. The stories are mine, not made up by my good friend Claude. The experiences are mine. The rants are most definitely mine.
AI helps me do the work. It doesn’t do the work instead of me.
Doping or carbon shoes?
So is AI the doping of the creative world? It’s a tempting comparison. A writer gets a little chemical help, wins the prize and the rest of the field never stood a chance.
But doping is cheating for a simple reason. Everyone agreed to the rules beforehand, and you broke them in secret. Nobody has written the rules for AI yet.
A better comparison is the carbon-plated running shoe. When those arrived, records fell and the purists cried foul. Athletics didn’t ban them. It set limits, and now nearly every elite runner wears a pair. A few people still grumble. The races go on.
For me, the line is simple. Using AI isn’t cheating. Saying you don’t use it when you do—that’s cheating.
Can you tell?
Walk round any toy fair and you’ll see packaging and characters that feel AI-generated. Too smooth. Eyes not quite right. Hands with a mind of their own. And a strange sameness, as if every character was designed by a committee that has never met a child.
When an amateur uses AI, it’s obvious. When a trained graphic designer uses it as one tool among many, I doubt any of us could tell.
One of my factories tells me openly that it uses AI for packaging. The others are more cautious. My guess is they all do.
I see the same thing on LinkedIn. Some comments on my posts have clearly gone through a chatbot, and they miss the joke completely. My humour is dry, a bit sarcastic, French filtered through twenty-odd years in England. AI doesn’t get it. To be fair, some humans don’t either.
Does any of this hurt the toy? Probably not. A child doesn’t care who drew the box. The problem starts when the box shows the toy doing things it will never do in real life. That was a problem long before AI. AI just makes it cheaper and more convincing. And a disappointed child means a returned toy and a one-star review.
AI in the toy box
Then there’s AI inside the toy itself.
I’m looking at AI-powered companion pet toys at the moment. But I have one rule: no internet connection. An AI toy that works offline, I’ll sell. One that sends a child’s voice to a server somewhere, I won’t. Not for now, anyway.
Parents are right to worry about where that data goes.
So are regulators.
Until the industry can answer those questions properly, the offline toy is the one I’m comfortable putting my name to.
Who wins, who loses
Let’s not pretend. Illustrators and packaging designers are losing work. That’s real, and it hurts people who are very good at what they do.
Small brands win. A start-up with a good idea and no budget can now produce packaging, a catalogue and a website that would have cost a fortune five years ago.
Businesses like mine win quietly, behind the scenes. I use AI to clean our CRM database, translate and build reports I could never have afforded to commission. I’ve even vibe-coded a few. I’m not a programmer. The reports work anyway.
Where this ends
Should packaging say “made with AI”? If regulators make it mandatory, I think that’s fair. Same rule for everyone, like the carbon shoes.
I’m optimistic. Every new tool in this industry has had its panic, and every time a balance was found. This will be no different.
Always forward.
For the record: I wrote this article. AI checked my facts, polished my grammar and, as you read at the start, corrected my memory.
Thierry is a seasoned international sales and marketing expert who founded Konomocha in 2015, has over 20 years of experience and helps toy and stationery brands expand into EMEA and APAC markets.
This article also appeared in Edition 55 of The Bugg Report Magazine
