Facebook Messenger's AI generated stickers are already proving problematic | indy100

Facebook Messenger's AI generated stickers are already proving problematic

Fox - 26 Houston / VideoElephant

Just a week after Meta's announcement of its new AI products things already seem to be going wrong.

The new AI-generated stickers that are available on Facebook, Instagram and WhatsApp seemingly have no filters when it comes to creating questionable and controversial content. Users have been sharing what the AI has allowed them create such children characters like Mickey Mouse smoking marijuana or Winnie the Pooh holding a rifle.

Artist Pier-Oliver Desbiens posted various stickers he managed to create using Meta's new AI tool saying "I don't think anyone involved has thought anything through". He attached screenshots of a child soldier, Karl Marx with breasts, the Pope holding a rifle, Elmo with a knife, and a pregnant Shrek, amongst others.

found out that facebook messenger has ai generated stickers now and I don't think anyone involved has thought anything through
— podesbiens.bsky.social (@podesbiens.bsky.social) 1696369572

Meta AI is currently in beta and available only the US, so it seems users have taken the opportunity to point out the large oversight they've made when it comes to the perhaps too wide a range of possibilities available with the sticker-generator.

One user wrote: "Got a feeling that Meta is about to be responsible for a whole bunch of AI-related US case law".

Got a feeling that Meta is about to be responsible for a whole bunch of AI-related US case law
— Rami Ismail (\u0631\u0627\u0645\u064a) (@Rami Ismail (\u0631\u0627\u0645\u064a)) 1696399708

However, some saw the hilarity in the situation, with one user saying they were going to enjoy the feature "until Meta gets sued into oblivion for copyright infringement".

Gonna have fun with this until Meta gets sued into oblivion for copyright infringement
— Baby Cheetah (Marxist) (@Baby Cheetah (Marxist)) 1696381154

Whilst others joked that they were now "pro-AI":

I am now pro-AI
— Jess O'Thomson \u267f\ufe0f \ud83c\udff3\ufe0f\u200d\u26a7\ufe0f (@Jess O'Thomson \u267f\ufe0f \ud83c\udff3\ufe0f\u200d\u26a7\ufe0f) 1696378611

i\u2019ve changed my mind, ai is good actually
— metal.txt (@metal.txt) 1696378310

i am saddened to report that facebook is now good
— Jesse Graham (@Jesse Graham) 1696380892

Meta's website features a blog post titled "Building Generative AI Features Responsibly", where they write "As with all generative AI systems, the models could return inaccurate or inappropriate outputs. We'll continue to improve these features as they evolve and more people share their feedback." The message suggests that Meta is probably going to get on top of this quickly so beloved children's characters are no longer holding weapons.

indy100 has reached out to Meta for comment.

Sign upto our free Indy100 weekly newsletter

Have your say in our news democracy. Click the upvote icon at the top of the page to help raise this article through the indy100 rankings.