Rogue AI “Regulation” – A Potential Worst Case Scenario

Authored by Kit Knightly via OffGuardian,

We’re now several weeks into our new global fear narrative – “Rogue AI is going to kill us all”.

The problem is stated, the reaction stage-managed, and we’re heading for the “solution”, which will inevitably be some kind of “regulation”.

They have been diligently vague about what this “regulation” might entail, and that means only one of two things –

1. either they don’t really know yet and are keeping their options open,

2. or they know it will be so unpopular they can’t afford to say it out loud yet.

We might be due some hints – maybe even more than hints – with the United Nations General Assembly beginning yesterday, and a state meeting between China’s Xi Jinping and Donald Trump later this week.

We can be fairly sure however that anything announced or posited this week will only be the beginning of what is very likely to be longer-term process. We’ve seen this already with regulation of social media and the internet in general. Multi-layered legislation forming a global patchwork of tyranny – that’s the model, although, with the effort clearly put into this “scary AI” storyline, I would expect them to move somewhat faster than the social media age verification laws, for example.

I’ve seen a lot of people suggest this will come down to securing a monopoly for the established AI giants, but, while this may well be a factor, I don’t see that as a sole end worth the means at this stage. Global media synchronisation is neither easy nor cheap, and to organise the biggest example of it since the dawn of Covid would suggest something more than stock prices are on the agenda.

As for real specifics – what any new rules might actually say – I’ve been working on my own ideas, and I wanted to share this shameless supposition with you all.

What might we expect from ‘anti-AI’ legislation?

Firstly, and most obviously, AI will no longer be free. No free access to generative AI, and potentially no more anonymous access to generative AI. It may well be that in the future if you want to use AI you will have to pay and/or verify your ID.

Almost certainly, they will make it so prompts are recorded, scanned and “potentially harmful” ones are forwarded to the government.

“What are you making with AI that you’re so ashamed of?” will be parroted line in response to any objection.

They will tack on stuff about AI being used to make “child sexual abuse material”, misogynistic or racist content, spread “hate”, or “digital violence” or whatever other buzzwords they need to win over hysterics from the left and right and justify this clause.

They have, in fact, already started doing that. A lot.

AI and anonymity fuel surge in digital violence against women
United Nations Report

Secondly, no more open source. If you make a model you will need to register it with the government and pay a registration fee, and have it inspected by “experts” who will shut it down if it’s not deemed safe.

Precedent for this already exists in the shape of the “Social Media Licensing Bill” currently passing through the UK’s Parliament, which demands anyone who runs a social media platform require a license to operate it.

At a stroke, these two moves would embed the AI giants monopoly, end in-house or open-source AI models, and cripple small independent businesses who increasingly rely on generative AI to reduce man-hours for simple tasks like editing, checking code or making images.

Third, we’re going to be told we need to protect ourselves from “rogue AI agents” masquerading as humans – so no more anonymous internet usage. If you want to open an account – for anything, anywhere – you’ll need “proof of personhood”.

That’s also been a talking point for a few months now.

We Need A Way To Prove Personhood Online – The growing number of AI agents roaming the internet will eventually force us to verify what the old web mostly presumed: that there is a morally and legally accountable person somewhere in the chain.
Noema Magazine

Your digital “Proof of Personhood” (they’ll probably brand it with a happy logo and fun little acronym like “proof of personhood and identity” or “POPI”) will be required to open any social media accounts, make online payments, buy hosting services…basically everything.

Little cartoons of Popi will teach you how to use your new proof of personhood to book a medical appointment or pay your taxes while protecting your kids from being killed by Skynet.

“Rogue AIs are hiding behind anonymous online activity”, they’ll say. “Proof of humanity is a small price to pay to prevent the apocalypse”.

That’s why the “Rogue AI” script has been so careful to include reports of AIs pretending to be humans.

They’ll likely go after every even partially unwatched corner of the digital world, constantly breaking new ‘revelations’ about AI agents hiding behind VPNs or using crypto currency to force an end to anonymity there as well.

None of these are especially “out there” suppositions. It’s rather like predicting a man will hit the ground after you watch him fall off a ladder. In fact we’re basically already there.

But is that the end of it?

Given the amount of hysteria being generated, it’s got to be possible these measures will be just the beginning.

It’s possible we’re going to end up in a deeper and darker place than that.

A “post-internet future” perhaps? Or at least, post ‘the internet as we know and understand it‘.

A future where internet usage and AI compute is limited by bandwidth, tokenised or even rationed.

The narrative would be quite simple –

“Excessive digital traffic and AI usage is bad for the planet and risking an AI-related apocalypse, so we must transform the way we use the internet”.

And unless we feel too skeptical about this step, let’s remember a lot of the pieces required to spin that fiction are already in play.

Firstly, they have established in a lot of people’s minds – through repeated propaganda and, frankly, tortured statistics – that data centres, ALL of them – are uniformly and irredeamably bad.

They use too much water. They use too much electricity. They release forever chemicals. They are worsening climate change. This is all stuff that has already been said, and with highly variable amounts of accuracy.

The motive behind that propaganda drive has never been completely obvious, but maybe it’s becoming so now?

After all, having established that all data centres – no matter where they are and what they are used for – are bad, it wouldn’t be a huge step to get to “it’s time to lessen our reliance on them”.

That means propaganda campaigns aimed at reducing the world’s internet usage – or rather, reducing ordinary’s people’s internet usage. And if not reduce it, at least make it more expensive.

“The era of infinite data is over”, the press may say, “people need to adjust to more realistic prices for internet access.”

Other articles will bemoan how “spoiled” we have become as a society, with things being easily accessible at the touch of a button, and that such a system was never “sustainable”. Inter-generational division will be encouraged, just as resentment of boomers has been so deeply and irrationally ingrained.

Clearly, they’ll say, a certain amount of internet usage is required for the world to function – emergency services, healthcare, the military, air traffic controls, taxation records and other government offices. These will all be deemed necessary and be officially labelled in corporate speak, SBDC – “societal-baseline digital communication”. Something like that.

Those services may have their own data centres set aside – what they’ll call “essential digital infrastructure”.

However, a massive percentage of our data centre usage is of course due to what they will call “non-vital internet traffic”, or some similar phrase.

This is social media posts, watching Netflix, online shopping, playing multiplayer games…that kind of thing.

However YOU use the internet in fact.

Your “non-vital internet traffic” could become a counter, linked to your digital ID (oops, sorry, “proof of personhood”) – like your vaccination records or your carbon footprint. They’ll probably even call it a “digital carbon footprint” in asinine explainer articles.

Vox or Salon will run listicles detailing “Top 10 Ways You Can Reduce Your Data Centre Use”, while the Guardian wheels out a bob-cut, glasses-wearing humanities graduate named Chloe to go into painful detail about how much her life improved now that she no longer lives online.

“Shopping in person, instead of ordering groceries online, helped me reconnect with my neighbourhood and myself”

“Watching old DVDs with my kids to save our internet footprint taught me to appreciate the little things”

The price of streaming services will either increase, or your watch time will be limited – at first they’ll probably let you choose which. The cost of “unlimited” data and internet services will go up as well, with “data centre friendly” bandwidth limitations being preferred.

These limitations won’t apply to uploads from your GPS-enabled devices, smart air quality monitors or smart water and energy meters, as these uploads will be classified as ICMD -“important climate-monitoring data”.

You’ll suddenly be encouraged to buy physical media again because it’s more reliable, and to shop locally, rather than online, because it supports small businesses. And this marketing will work because it’s largely true.

But don’t think there won’t be a catch. eg – the new physical media will have a new green label on the box that says DCN – “data centre neutral”., and this tiny addition will cause the prices to go up exponentially, but of course that will be a small price to pay to prevent the apocalypse, too.

Oh and the “new” generation of physical media players won’t be backwards compatible with your existing DVDs or BlueRays – so you’ll need to purchase everything again, at the new inflated price, and with availability strictly limited to movies and music that doesn’t offend current sensibilities.

Of course, you realise none of this will really be about avoiding any apocalypse.

It will be about compartmentalisation. Cutting lines of communication.

Running small business or personal websites in general will become more expensive as the “data centre tax” is added to hosting providers.

Much like the “sugar tax” on fizzy drinks and snacks in the UK, this tax will notionally be paid by the hosting providers, but will be passed onto the customers via increased prices.

Major corporations will be able to easily afford these increases, but smaller independent sites will feel the pinch.

It will be cheaper to pay corporate resellers to cohost, but then you’ll have to sign up to agreements about not publishing “misinformation” and, again, provide your proof of humanity.

If you run a website your data centre usage tax will be 0.04 per local visitor, but 0.10 per international visitor – since serving data further is more compute heavy.

You’ll be able to get reductions in hosting costs if you use a “local only” server, accessible only to those resident in your home country/region/bloc etc.

Automatic translation, we’ll be told, is very heavy on AI usage. So it will be increasingly cheaper to only publish in your native language too.

It won’t matter if any of this is true or not, they’ll just say it, as they do.

On social media, the more followers you have the more data you’re using, so after a certain point you will have to either pay to maintain your followers, or pay per post to make sure your followers can see it.

Of course these limitations won’t apply to legacy outlets like the BBC, CNN or their ilk of course. They will be exempt, because they are part of the “essential information distribution network”.

And all the time, the fear of “rogue AI” will be there, to be picked up as convenient. A digital pandemic that, like Covid, can be started and stopped on a dime by the digital version of “testing protocols”.

Kill switches and “digital firebreaks”, allegedly installed to quarantine “rogue” AI agents, will enable swift isolation of any anti-establishment opinions or information that breaks through to an audience of any size.

An independent news YouTube channel getting a lot of views? Oh no, it’s been reported to be AI and has been suspended pending investigation!

An alternate media website is getting a lot of views? Oh no, the server it’s stored on has been compromised by AI and needs to be temporarily isolated!

Again, obviously not one word of this needs to be true. They will just say it. Like they just say everything now.

And the real kicker is it won’t matter if global internet traffic does decrease or data centre use really declines. They probably won’t even check. The real numbers don’t matter.

Does this seem fantastical? Maybe. But we have seen similar before, with both Covid and climate change. The ‘solutions’ don’t need to work, because the ‘problem’ never existed. It was just a narrative constructed for a given end. All they need to do to ‘prove’ they were successful is buy some scientists to publish doctored papers that torture numbers and spin lies – and that will be enough.

To be clear, I’m not saying all of this is definitely going to happen, in fact I’m not sure all of it is even feasible. But it’s what I think they would like to do if they can, and may try to do.

We’re too far down the road to Dystopia to be complacent about the intentions of our keepers. And feasibility has never been a barrier to the ambition of tyrants to be tyrannical. We had years of attempts to “ban end-to-end encryption”, despite it essentially being impossible.

Anyhow, whether these worst case scenarios materialise or not, the direction of travel is the same: toward less freedom, less choice, less privacy. Same direction as every other major narrative drive for the last ten years if not longer.

More specifically in this instance it could be about turning the internet from global unlimited access to local pay-as-you-go.

The current internet, even with all its restrictions, has been instrumental in fomenting opposition to globalist policies, distributing evidence of false flag attacks, open source investigations into corruption in the elites – it is the greatest tool for the democratisation of information ever invented, and the people running the world have ALWAYS hated that aspect of it, even while they exploit the benefits gives them.

In the name of protecting us from the AI apocalypse and saving the planet from toxic data centres, I think the elite might sense a chance to regulate that particular genie back into its bottle.

Yes, as I said earlier, the second part of this is all supposition, a potential worst case scenario. Maybe I’m just getting paranoid and inventing another kind of fear porn!

I sincerely hope that’s all it is.

I will never be happier to be wrong.