When AI Gets to Play Judge, Jury, and Executioner
So a tech YouTuber with over 350,000 subscribers just had their entire account terminated by YouTube’s AI moderation system. No warning, no human review, just poof – years of work gone. And the kicker? Good luck getting a human at YouTube to even look at your appeal.
This isn’t just about one YouTuber having a bad day. It’s a perfect example of what happens when we hand over the keys to algorithms and call it efficiency.
The whole situation highlights something that’s been bothering me for years now. Tech companies have essentially created these digital fiefdoms where they make all the rules, and if you get kicked out, your options are basically “appeal to the same AI that banned you” or “go viral enough on social media that someone at the company notices.” That’s not a system – that’s a lottery.
Think about it: we’ve got regulations for utilities, banking, telecommunications. If Telstra cuts off your internet wrongly, there’s an ombudsman you can contact. If a bank makes an error with your account, there are established processes and actual humans you can speak to. But YouTube, Facebook, Instagram? They’re handling people’s livelihoods, their businesses, their creative output – and there’s essentially no recourse beyond hoping an algorithm might reverse itself.
Someone in the discussion pointed out that everything gets labelled as “AI” these days, and they’re not wrong. Half the time we’re probably just dealing with the same dodgy algorithmic moderation that’s been getting things wrong for years, but now companies can shrug and say “the AI did it” like that somehow absolves them of responsibility. It’s a convenient excuse, isn’t it? “Sorry, the machine made a mistake, nothing we can do about it.”
The really frustrating part is that this isn’t new. I’ve been working in IT and DevOps for long enough to know that automated systems make mistakes – that’s why we have monitoring, rollback procedures, and yes, humans in the loop. But tech platforms have decided that human oversight is too expensive. Never mind that someone’s entire income stream might depend on their channel. Never mind that false positives are inevitable with any automated system.
Last month, my daughter was trying to upload a video project for school to her YouTube channel, and it got flagged for copyright violation. The video contained exactly zero copyrighted material – it was all original footage she’d shot herself. The appeal process? Submit a form and wait. No explanation, no way to speak to anyone. Fortunately, it was just a school project, not her livelihood, but it gave me a glimpse into how kafkaesque this system really is.
Here’s what really gets me: the same tech companies that are rushing to replace customer service workers with AI chatbots would never dream of replacing their executives with algorithms. Someone joked in the discussion about looking forward to when AI takes the jobs of executives, and honestly? That hit home. Why is it always the workers and the users who bear the brunt of “efficiency measures” while the C-suite continues to collect millions?
We need proper regulations around this stuff. Real dispute resolution processes that don’t involve “get lucky on Twitter.” Requirements for human review when livelihoods are at stake. Actual accountability when automated systems get it wrong. The EU has started down this path with their Digital Services Act, and maybe that’s what it’ll take – regulatory pressure – because the tech platforms clearly won’t implement these protections voluntarily.
The environmental and social costs of AI are already concerning enough – the energy consumption, the job displacement, the reinforcement of biases. But when we’re also using it as an excuse to eliminate basic customer service and fair processes? That’s when I start thinking we’ve lost the plot entirely.
Look, I’m fascinated by AI technology. The progress we’ve made is genuinely incredible. But there’s a massive difference between using AI to help humans work better and using it as a replacement for basic fairness and accountability. Right now, we’re doing the latter, and people are paying the price.
Until we get proper regulations in place, the best advice seems to be: don’t put all your eggs in one platform’s basket, back up everything obsessively, and maybe start building your audience somewhere you have more control. It’s a rubbish situation, but it’s the reality we’re living in.
The tech industry built itself on the promise of connecting people and democratizing content creation. Somewhere along the way, we ended up with a system where an algorithm can destroy someone’s business overnight and there’s no one to call. That’s not progress – that’s just corporate efficiency dressed up as innovation.
We can do better than this. We should demand better than this.