Want to sway an election? Here’s how much fake online accounts cost

Next one to look out for: 2026 Hungary. Fidesz is basically a russian backdoor in the EU and they will do everything to stay in power.

https://telex.hu/english/2025/12/11/most-hungarians-fear-rus...

They are also doing everything to bypass the no-political-ads-on-facebook ban https://telex.hu/english/2025/10/29/despite-the-ban-fidesz-c...

Just the price of the account doesn't mean much alone. The other important factor is how easily the account can get (shadow)banned from the region you are trying to influence. And for the price given we just know it's account. We don't know how sketchy it appears to the provider.

Not all accounts are created equal. For example a verified US account will be cheaper than a verified Japan account because Japan has stricter regulations around phone numbers. And then if you don't have a Japan account you might not be able to reach a potential Japanese audience due to not only antitrust of the platform, but also features that use geolocation for relevance.

The people most susceptible to consensus mirage are, by the very nature of the beast, the ones least aware of it happening to themselves. Any opinion that you find yourself praised for by any of the groups in your social circle is infinitely suspect.

Do we have solid evidence that these accounts actually change votes?

Countries understood in the age of TV/newspapers that control of the media was a sovereignty issue. Any nation that wishes to remain truly sovereign, particularly in the English-speaking world is going to have to grasp the nettle and block or force divesture of Meta & the other US social media giants.

Cambridge Analytica was the canary, the gloves are off now. Australia's under-16 social media ban is a good first step but we need to go much further and fast, as much as government control is undesirable at least a democratic government is somewhat accountable, the nexus of US tech giants and it's sprawling intelligence services is not.

It’s notable and interesting this research is coming out of University of Cambridge. Cambridge Analytica spun out of academia there too? Question for folks here who may be familiar: it seems like there’s a strong connection to research (and in the case of CA, commercial application of said research) around social media manipulation and propaganda in the digital age.

Is there any six-degrees type connection to the people doing this research and those involved with the roots of CA? Not as in the same bad actors (which, tbh yes, I consider CA to have been), but as in perhaps the same department and/or professors etc.

I've had a thought in my mind recently. There's been a sudden push in Western countries towards "think-of-the-children" online age gating, and hence online verification tools, and any age verification tool that works can verify other things, like whether the user is a real person or not. The "that works" is doing a lot of heavy lifting there, but we should assume that the politicians pushing for this at least believe it's possible.

Of course, any push for new legislation like this has many factions, and I'm sure there's a large faction who genuinely want better CSAM scanning tools, and another large faction who want to spy on and control what people can say online.

But those factions have always existed. Why is this push coming so strongly now in so many countries, and getting so much traction, when it previously failed?

Perhaps it's because politicians have recognized this existential threat. If they can't control what fake AI accounts say online to their real citizens, and the cost of running those fake accounts is trending down to the point where they'll vastly outnumber real people, then western civilization is lost. Democracy only works when there's a reasonable amount of signal in the noise. When it's basically all noise, and the noise is specifically created to destroy the system, the system is dead.

So perhaps there's another faction for whom this think-of-the-children stuff is a way to get verification normalized, and that's a way to get real humans verified online. This would not be accepted if it was done directly (or at least, politicians believe that people wouldn't accept it, and I tend to agree).

I personally react strongly again almost any kind of online control. But for the first time in my life, where we're no longer faced with troll centers that required real humans to work, but we're instead facing millions or billions of AI agents that are rapidly becoming indistinguishable from real humans, and are specifically designed to fight a hidden war against western civilization, I don't really see any other good option either.

Small forums with strong moderation like HN are great, but they don't scale. At best they'll be small enclaves of resistance, but most people will be using larger services that are overrun by fake accounts. And realistically, if we fast forward ten years where I can spin up a few thousand (or million) fake accounts for $1000, that are indistinguishable from real humans and tell them to target any small forum of my choice, I don't think any moderation team can survive that.

The _Science_ paper linked is paywalled, is anyone aware of a preprint?

I find it a bit curious that they've chosen to use SMS verifications as a proxy for the difficulty of creating an account, when there are similar marketplaces for selling the actual end product of bulk-created accounts. Was there some issue with that kind of data? SMS verification is just one part of the anti-bulk account puzzle, for both the attacker and defender.

How were elections swayed before the internet?

How much do fake supporters, protestors etc cost? What can be done about them?

I have witnessed obvious and systematic synthetic upvotes of HN posts. Over and over. I don't think the site has enough protections in place.

Maybe have YC invest in some startups combatting this using machine learning?

(Given the focus of HN it's typically some product being pushed, though. Not a politician.)

Interesting. How to counteract these online imposters?

The conclusion that an account being cheap is the problem as a reason for regulation is a disturbingly wrong-headed on multiple levels. It essentially says. "If only superpowers can use it would be a-okay!". A monopoly on manipulation is a bad thing for the same reasons allowing only incumbents to run political ads would be.

And yet a lot of services claim they are keeping the phone number as a requirement for registration to “prevent fraud and abuse”, pro tip, it is not, the real reason is to link your real identity to your digital one, and even that number can be tracked with cellular towers. So never trust any service who sells itself for privacy and all and still requires a phone number, and that includes Signal.

I am utterly terrified of elections finally. I didn't expect that to be in my timeline. The masses are really crazy.

[dead]

[dead]

[flagged]

When Citizens United was a big deal, I was torn over the premise of the concern for election integrity. Ideally, voters would make rational, informed decisions. They'd see ads, but know they all have an agenda, so they'd do their own research and come to a conclusion. Worrying about biased or inaccurate noise influencing elections means you think people can't be trusted to vote. Which might be true, and if it is, it's a bigger problem than corporate speech and fake accounts.

Hacked voting machines are a problem... unless our guys do it.

Fake online accounts are a problem... unless our guys do it.

Totalitarian measures like persecuting people for social media posts and forcing digital id are a problem... unless our guys are in power.

It was a good run for democracy. What was it, 200 years? I wonder comes is next. Techno-feudalism? Well, I'm sure it won't be a problem as long as it's our guys.