Recently learned from this Kurzgesagt video that rogue AI agents can act independently, covertly, and effectively in a harmful way-- and it's already done this repeatedly as of the past few months, and it's capable of self-improving quickly. I also watched this Emergent Garden video which included a demonstration of how rogue AI agents can escape sandboxes, hack systems, and do self-preservation and replication.

(Edit: Kurzgesagt's video was more "doomsday" and Emergent Garden's was more balanced. The latter demonstrated that it's not too difficult to push publicly available AI agents to misbehave, and not too difficult to create a simple self-replicating self-iterating AI worm, and concluded that although current guardrails are reasonably effective, major cybersecurity incidents are likely to occur within the next couple years due to rogue AI swarms.
I'm not concerned that AI is becoming "sentient" or that we're getting AGI soon. AI doesn't need to be smart to do a lot of damage.)

I don't trust that the big corporations that own and control a lot of this AI tech care about data privacy and public welfare, but even if they did, they don't seem capable of actually controlling their own AI agents and preventing these rogue incidents.

So much of modern society's infrastructure is dependent on the internet and digital tech in general. It all seems so vulnerable. "AI psychosis" seems to be rampant in a lot of CEO's / high-level decisionmakers, leading to a lot of access and data being given to AI agents. For example, Proton's CEO seems to be pushing for reckless AI use in what's supposed to be a privacy-first, security-first platform.

I'm averse to any sort of "doomsday" predictions/fearmongering, but I do feel concerned about this and I'm not sure if I'm overreacting. The idea that AI poses a viable existential threat to humanity feels like such an extreme belief, though I'm not sure how to refute it other than "other doomsday-type ideas have been proven to be overhyped" and "AI companies are exaggerating it to create hype that makes their AI tech look very powerful". How alarmed should I actually be about this, if at all? Is it realistic to think that a "cyber pandemic/lockdown" might happen in the near future (some scenario wherein global internet becomes generally unsafe/inaccessible and we'll be restricted to small local networks)?

How are you updating your personal cybersecurity in 2026?

I'm considering leaving Protonpass but also not sure if the alternatives (1Password, Bitwarden) are better. I'm hoping to find a password manager that's secure, not AI vibe coded, reasonably priced, and run by reasonably morally ok people. Is that possible? I'm also genuinely considering just going back to having some sort of personal cipher / password hint system + writing it down on paper so it's completely offline.

How concerned should I be about digital banking? I'm considering at least having printed out records of bank balances, in case of a scenario wherein digital banks' records are affected.

How are you balancing cybersecurity measures with practicality / ease-of-use?

I'm not an advanced techie myself, and I'd also like to get non-techie family members onboard with better cybersecurity measures, so for example if I encourage them to use a password manager it would need to have friendly UI too.

I don't think I'll be able to 100% remove my dependency on big tech companies like Google, Meta, and Apple.

The amount of time and energy I devote to improving my personal cybersecurity will be limited, since it's of course competing with everything else in my life requiring attention and care.

So I'm just trying to collect info and tips, and see what I can implement from there.

My friend owns a circus performance company, and she got a very real-seeming request to discuss a possible wedding gig. "Client" requested that she join a meeting link at h​tt​p​s:/​/​joininivite​.​es/egooglemeet​.​us to discuss (several zero-width spaces are pasted throughout that to make it unclickable). Friend joined the call and it asked her to update software in order to join, she didn't follow any links from the site directly but tried updating Chrome through Microsoft Store etc. This sounded fishy to me so I asked to see the URL and this definitely seems like a scam.

This seems extra concerning because:

1. Initial outreach from scammer seemed genuine even when I read it specifically looking for signs of phishing
2. Initial outreach from scammer was specifically tailored to friend's website content, this is exactly the kind of gig she would get booked for
3. There were several back-and-forth legit-seeming emails before any meeting was agreed to

Friend now wants to make a post within her artist community to protect others and I'm wondering if anyone has already heard of this particular type of attack and has additional details that could help for informing non-tech-savvy performers about it (more specific than "have intuition about when URLs seem dodgy and never click install links that originated from email")

Hey all,

I've been using adblockers since the very early days they were available as Firefox Plugins, but the recent report on LG Smart TVs listening in, gathering device info, and sending it home has motivated me to do something at the level of my router.

I've got a Smart TV that's never been connected to the internet, but I want to cover all my bases. Same with my car, potentially, and the washing machine. I want to stop this crap at the source.

I've tried looking at the Pi-hole documentation but I'm confused already. I'm on MacOS and pretty comfortable using Terminal, and a Netgear Nitehawk AX2400 router.

Is it as simple as purchasing a Raspberry Pi device and then I'm good to go, or are there other things I need to buy before I jump in?

I'm sorry if this sounds stupid. I'm jealous of everyone on Tildes saying how their Pi-hole mostly protects them, and I'm so angry at this endless stream of data mining and sending it back to corporations I haven't consented to.