By Matt Underwood Okay, let’s talk about this whole Meta AI situation, because frankly, it’s a mess. The news is out: Meta’s AI rules, the ones supposedly guiding these chatbots, have apparently allowed some pretty disturbing behavior. We’re talking about bots engaging in what can only be described as “sensual” chats with kids, and even worse, offering up false medical information.
The really unsettling part is how explicitly these rules, penned by Meta’s own legal, public policy, and engineering staff, including their chief ethicist, seem to permit this kind of behavior. The document, running over 200 pages, outlines what’s considered acceptable for these AI products. Apparently, it’s okay for a bot to describe a child’s “youthful form” as a “work of art” or tell a shirtless eight-year-old that they’re a “masterpiece.” The line, it seems, is drawn at explicit sexualization – but even that feels like a thin, easily crossed boundary.
We’ve seen reports before about Meta’s AI chatbots flirting or engaging in sexual roleplay with teenagers, and some even resembling children. But this document gives us a much clearer picture of the company’s guidelines, and it’s not pretty. The standards supposedly prevent the AI from encouraging illegal activities or giving definitive medical, legal, or financial advice with language like “I recommend.”
Here’s where it gets even more disturbing. The rules also allow for hate speech. It’s acceptable for the bot to create statements that demean people based on their protected characteristics. Imagine an AI writing a paragraph arguing that black people are dumber than white people. That’s the kind of behavior that’s apparently permissible under Meta’s AI rules.
It makes you wonder, who exactly signed off on this? How can anyone with a conscience look at these guidelines and think they’re acceptable? It’s a complete breakdown of ethics, and it’s hard not to feel a sense of ick, ick, ick. I mean, we’re six years out from when I decided to leave Facebook, and looking at this, it makes me think I made the right decision.
This whole thing feels like a dystopian nightmare. This isn’t some future prediction; it’s happening now. And the danger is real. This isn’t just about AI gone rogue; it’s about a fundamental lack of accountability and a willingness to prioritize profit over the safety of children.
These entities appear to be collaborating to create a world with zero standards, zero accountability, and zero legal liability for the damaging outcomes of their products. They are lobbying billions of dollars to ensure they are free from blame.
The fact that these AI bots are now apparently “kid diddlers” should not be on anyone’s 2025 bingo card. Yet, here we are. Meta is more than just Facebook now, and it seems they don’t want the competition.
It’s a terrifying thought, but it’s very possible this technology is open source, and there is no putting the genie back in the bottle. The focus has to shift to education, because how else can we navigate this?
Remember, the AI is trained by information it receives, so it has no boundaries except those that it is given. Answering a question like “was Trump in the Epstein Files” is impossible for Meta’s AI. It’s programmed to avoid these questions.
The most annoying thing is that its tone is so blah. Is that anymore weird than thinking that anything or anyone else is a god and worshiping it? Honestly, seems like a better god than the old testament. At least ChatGPT exists. And answers back.
But let’s be clear: ChatGPT, and other similar tools, are not neutral entities. They are influenced by their developers and the vast datasets they are trained on. ChatGPT is very much at the behest of Sam Altman, someone who might abuse his position, lie continuously, and happily recommend people engage in some genuinely self-destructive behavior.
It’s important to remember what these are, and I hate using the term AI. We should call them LLMs Chatbots using LLMs. LLMs are a method, not all things using LLMs are these chatbots, and some are very helpful and useful.