Profile at Risk: Facebook’s AI Security Needs Restriction

The Downfall Of One Is The Uprising Of Another

Facebook, a social media platform that has been around since 2004 and used by 2.11 billion users daily, is what we would call a household name. It singlehandedly took out Myspace, a beloved platform for those of us that were around to use it, it was unmistakably the best around. We had music when you clicked on our profile. We had a list of top friends. We weren’t being banned left and right for a cuss word. Though Myspace is still active today, it peaked right after 2008, 5 years after its launch. Facebook dealt it a deadly blow with its 100 million daily users by the end of that same year, 2008. “The Book” has grown exponentially with not only its userbase, but its stocks and fortune, bringing in over $200 billion ( over $60 billion of that is profit ). They have approximately 75,000 employees with the number of employees peaking in 2022 with 86,482. Though they have slowly laid off their employees since 2022, they have dove further into the AI scene.

Facebook has driven it’s AI to take the reigns of many aspects of it’s multi billion dollar platform. One being, security. AI security, though cost effective, doesn’t seem like something we should have with freedom of speech. But, freedom of speech stops at Facebook’s front door. FB being a private company, the constitution is not honored or enforced. There are numerous things that Facebook chooses to not enforce and then there’s the the “problems” they choose to enforce, strictly.

The Security System Has Four Enforcement Mechanics

• Regulation and Limiting – The AI system can be triggered by viral misinformation or overuse of AI. (sounds familiar)
• Throttling and Suppression – Also called a “shadow ban”, the AI security team can be triggered by offensive/hateful remarks or similar statements and/or words.
• Punishment and Removal – Also known as, “Facebook Jail”. This is for the bad guys. The ones who try to sell drugs, illicit materials, they run scams, harassment and hate speech.
• Banning – Repeated violations and “direct triggers”, such as a phrase you used before and they stripped you of your privileges.

Facebook and Its Tactics

Sudden drops in followers, reactions/engagements and views are blamed on algorithm shifts, feed updates and automated flags. Automated flags in which you cannot auto un flag. Being shadow banned “throttles” your profile to where your reach is very short. Most of your followers and friends will not see your posts. They may even wonder if you’re still their friend, or what happened to you. The shadow ban is very real.

When you post on Facebook, whether it’s a picture, video, phrase, quote, music lyrics, Bible verse or whatever else it may be, it is sent into a virtual funnel system. Through highly advanced language models. The system does not flag based on a set in stone list of bad words. As the post goes through the funnel, it automatically picks up the worst of the worst, “direct triggers”, such as terroristic threats, self harm/mutilation, suicide and the rest you can probably come to conclusion on. Once the worst of the worst, or the easiest of phrases, words, pictures and videos are separated from the rest, the “good” posts are sent on their way to the next steps. Picking up on a few other words and phrases such as hate speech, harassment, drugs, illicit materials and scams. Some of these slip through with the rest, though. The AI system begins to fail, here. As it fails, the posts are sent further into the funnel and more phrases, pictures and words are taken out. Flagging posters from all over the world. Pretty soon, what comes out of the other side is posted without problem, or you might think.

When one of your posts bring forth the AI centurions with their flags staking claim on misuse and outright lawlessness, it brings with it the punishment phase. First timer? They may go easy on you with a short sentence, with a slap on the wrist. Even if you feel your post is in the right, it’s a flag. Want to fight the opinion of the AI software that Zuckerberg so diligently built into his cash cow? Good luck. When you appeal a claim made by Facebook security for the blatant disregard for its regulations and rules, it is sent to an appeal team of sorts. Get this, there are so many appeals that it is sent to an endless AI automated appeals loop. “They” review your appeal and will quickly send you a message letting you know you have lost the battle. Their aliens know better than you and are smarter. Joking aside, it’s tiring trying to go against a machine as large as Facebook. I have had my fair share of battles with them and I have only won one appeal out of many.

Once you’re busted, you are now on a watch list. Not that kind of watch list, but their watchlist. A platform wide list of potential problematic users. We now have a “lower trust score” than the others. Strange, huh? It’s like a credit score in the Meta world. Our words now carry so much weight that we literally lose brownie points with robots. Thanks, scientists. Once we have a restriction on our profile, we are more apt to another one… and another one. We are quicker to be “judged and sentenced”. We move to the front of the line, or the back, depending on how you view it. We carry negative history for up to a year. Yes, you mess up in the Meta verse and suddenly you’re a nobody for up to a year. You don’t have the credit score that your Aunt Barbara does for posting only puppy dog pictures and the occasional questionable political meme. Isn’t it strange that you could say one wrong word and it completely screw up your business? Imagine Facebook is one of the platforms you rely on for content creation. It has your audience, they’re funding your dream and then BAM! Facebook giveth and it taketh away. It doesn’t stop there, though. We now have stricter AI scrutiny and you can now be flagged for minor/borderline words and phrases. Even if they aren’t categorized as hate or harassment, terroristic or self harm. You now can be punished for simply pissing off the wrong AI security member. Rufus the AI bot has a hard day at home and now he’s clocked in for the night shift and suddenly, you, of all people, log onto Facebook at 10 o’clock P.M. on an AI school night and write some off the wall comment on a news article and Rufus has had enough. You are now at his mercy and he says the comment has to go and you’ve been warned. Now, you’re upset, you appeal it. Go on about your night and Rufus says, nah. You are wrong and you cannot post again. Not in groups. Not in a live stream. Not even in a phone call. Rufus is a problem.

The Red Flag of Doom Presented By Rufus.

How To Win Rufus Back

So, you’ve exhausted your options and are now at the mercy of an AI security team member named, Rufus. My, my how did we get here? You can win Rufus back with a few options once you’ve been restricted. When Facebook decides to let its minions lift your restrictions/ban, you have to build trust, again. Gain your “loyalty team member points” back. You literally have to act human. You cannot post over and over, too quickly. You cannot react too many times to too many posts. You have to act human in order to win back a team of AI misfits. Post clean content and cool off for 24-72 hours. Man, they run a tight ship!

There Is Hope!

We can talk to a human? We can contact a human? Yes! Only a couple of ways though. Pay for Facebook! Yes, you can become “verified” and it falls under paid support. By getting a little verified mark next to your name, you now have the status symbol of the elite and the AI security backs off. Humans come out and ask, “is that you? a real life human!? why, we haven’t seen one in YEARS!”

The only other way? When you appeal your AI conviction, you can have a final appeal with Facebooks notorious “oversight board”. So, is there hope? Not really, but at least we’re not totally doomed without human connection and contact. Facebook may be the worlds leading social media platform, but it is far from superior.

Zuckerberg, #4 On Forbes Real-Time Billionaires List With A Net Worth Of $266.4 Billion.

Good luck out there, team. Watch out for the giant red flag of an AI bot.