"EU staff banned from using AI-generated content in official communications" Can confirm - we don't use AI generated images on our socials for instance.

[hpod16@eupolicy.social]

@_elena
It's not exactly a new thing, it's just a recent politico article picking up on the Commission's staff guidelines that have been around for a while

[hpod16@eupolicy.social]

@petrikas
sometimes stuff like this slips through the cracks. In this case it's a photo montage with images taken from a stock photo portal, and the initial images were probably generated by AI and not labelled as such... We've become a lot more vigilant to this kind of thing since, and organised trainings for designers on AI, so they can better spot this kind of thing and understand why it's best to avoid.

[petrikas@mastodon.art]

@hpod16 This is excellent news altogether. Thank you!

[hpod16@eupolicy.social]

@nihkeys
No, I agree. The headline is nice and catchy, but the article reeks of commercial interests of the AI business people.

[_ryekdarkener_@mastodon.social]

@hpod16

However the next step is to do a scanning for sense in the handmade communication.

[no_brainer@mastodon.social]

@hpod16 „Can confirm - we don't use AI generated *images* on our socials for instance.“

And how about the rest of the content? https://getmona.app/rich_text/172183068007

[faithfulljohn@mastodon.scot]

@hpod16

[brekke@eupolicy.social]

> The risk that deepfakes could erode trust “should not paralyze you into doing nothing,” said Renout Van Zandycke, who runs the Belgian communications agency Exposure, which has advised several political parties. “Doing nothing is equally as bad.”

So spreading disinformation is only as bad as not doing it. Got it. Excellent advice.
@nihkeys @hpod16

[ladupla@mastodon.social]

@hpod16 Okay

[richlv@mastodon.social]

@hpod16 @nihkeys Right?
Avoiding slop isn’t «doing nothing».

«Shoplifting isn’t great, but it’s better than doing nothing!»

[hpod16@eupolicy.social]

@richlv @nihkeys
Also, there's no ban on experimenting and educating on AI. The commission isn't doing nothing in terms of trainings etc...

[hpod16@eupolicy.social]

@brekke @nihkeys ‍️

[hpod16@eupolicy.social]

@no_brainer
We also don't copy paste AI-generated text onto our socials if that's what you're getting at. Our content creators are pretty passionate writers, I don't think any of us are keen on outsourcing the creative part of our work to AI.

[veza85ue@eupolicy.social]

@richlv @hpod16 @nihkeys Well, yes because #PoliticoIsTrash

They're owned by Axel Springer and run by Mathias Döpfner. I don't know what their latest relationship status is with their former (?) investor Peter Thiel (they had a falling out at some point), but those names should give you an idea of what kind of coverage of EU matters in general and EU regulations on slop in particular you can expect. And yet, the Brussels bubble treats it as its "newspaper" of record.

https://eupolicy.social/@Veza85UE/113782773270529645

[no_brainer@mastodon.social]

@hpod16 Great! Love that!
Thanks!

[tobiwankenobi@kolektiva.social]

@hpod16

In other words, it's okay so long as it's used internally, where no one can see, huh?

[technically_good@techhub.social]

@hpod16 Yes!!

[stevo618@mastodon.social]

@hpod16 Politicians here in the U.S. don’t want people knowing the truth. Many of them don’t know the truth themselves!

[nihkeys@mastodontti.fi]

@Veza85UE Thanks for this. I wasn't aware of the background. Also I think surviving being a "tech journalist" nowadays will likely require some kind of buy-in into the "AI is great"-narrative.

@richlv @hpod16

[veza85ue@eupolicy.social]

@nihkeys @richlv My pleasure. I used to have an entire folder of receipts because friends don't let friends click on #PoliticoIsTrash One of the all-time classics:

https://eupolicy.social/@Veza85UE/111487567167710407

@hpod16