Meta's Queer BIPOC AI Profiles Backfire. Again.

Yep, you read that right. The Internet may as well be dead.

Today's post will be short and sweet, highlighting what happened and why it's such a grand showcase of unfettered AI bias and a point towards meaningful AI ethics laws and guidelines. Let's dive into the first real post of the year!

So, AI profiles?

Meta, the company behind Facebook and Instagram has begun to delete their AI profiles amidst severe backlash from users over the AI slop these profiles have been posting. Take the posts from user / AI profile @hellograndpabrian as a taste of the riveting content they're posting:

You might be wondering: wait, Meta was making AI profiles? It was! Some four or five days ago, articles started making the rounds about Connor Hayes, VP of AI at Meta, discussing their "big bets" regarding AI bots on Facebook and Instagram.

Since then, Meta has supposedly started walking these AI bots back. Fun fact: this isn't the first time that Meta has attempted (and walked back!) these profiles on their platforms. According to the Guardian, Meta attempted something similar in September 2023 and "killed them off" around Summer of 2024. It really begs the question, what are they trying to accomplish?

The idea is that Meta wants users to be as engaged with their platform as possible. If you start receiving words of encouragement from "random strangers," and see posts that align with your views, then you're more likely to engage. Alternatively, content that makes you upset will also push engagement. Influencing you towards a certain lifestyle under the guise of an internet word-of-mouth recommendation from said "random strangers" may be a way of selling you a sponsored product or service. This, effectively, lets Meta control what information you see, down to the pages and profiles you're interacting with on a daily basis. Nothing could possibly go wrong with that.

Cloud-powered Stereotypes

What's most notable about these AI profiles is just how straight up biased they are. The example that comes up most to mind is their AI puppet Liv, a Meta proclaimed “Proud black queer momma of 2 & truth-teller.” (Please note that Liv's profile has since been taken down on Instagram)

Liv, according to her own testimony, was not made by any people of color! Even Liv can see the issues with making an AI with an identity nobody on the team actually identifies with. At the very least, she holds true to being a truth-teller.

From a personal perspective, if "Liv" was a real person that was black, queer, a mother of two, and willing to speak up about issues that matter, she'd be a queen and an absolute amazing person in my book. It leaves a void in my heart where instead of a real person, it's a profile fabricated by Meta to co-opt these identities for financial gain.

Meta has also been slammed in the past for suppressing queer content by several outlets. Here's a great post from Tucker Casey on LinkedIn that encapsulates some recent concerns:

How about dating advice?

The other example I'd like to bring to light is the profile of AI figurehead Carter on Facebook (also deleted).

Carter was marketed as a relationship coach. This also creates a sort of situation where Meta and their AI profiles can persuade you towards taking a certain action. Is your wife mad? Maybe get her a nice Pandora bracelet. Having trouble connecting with your partner? Surely a Meta-approved couple's counseling practice will solve the problem. It could even get more insidious than that, but I'd rather not get too deep into it for my own sanity.

Conclusion

I don't think Meta is going to pump the brakes on these AI profiles anytime soon. What is most important in this is that these profiles were easily identified and subsequently shamed into oblivion, for now. But what if these profiles coerced someone to do something untoward? Surely, AI has never, ever given bad advice.

Meta will also not be the last company to implement something like this, they're just the latest being caught by users. Being aware of these types of AI profiles and protecting your loved ones from being affected by what is essentially Meta propaganda is critical to navigating the digital age. Don't want to have to deal with issues like this? Consider moving social media sites to one that doesn't stuff AI stereotype profiles down your throat.

I realized that most of my posts of Cha Cha are of him sleeping. I swear he has energy, he's just a lazy boy.