• The phone number requirement
• Is threat modeling dead?
• Threat modeling isn’t dead
• Learning how to threat model
• Hiding in plain sight was always the best strategy
• So, what does that look like for the average user?

I’ve been using Signal for as long as I’ve been a privacy researcher and accelerationist (fancy word for “passionate advocate”). Over those years, I’ve heard so much FUD and outright nonsense about popular privacy tools, but for some reason, Signal always seems to get the harshest attacks.

§The phone number requirement

First off, I can understand the concerns over this… at a surface level. Once you dig into it, it’s not really that big of deal, though. Signal guards most metadata and Sealed Sender is the cherry on top. It ensures that unless both you and your contact are under active surveillance, it’s extremely difficult to figure out who you’re talking to, or who is talking to you. If you and your contact are both under active surveillance, hiding metadata really doesn’t change much at all, no matter what chat app you’re using. There’s also the practical aspect: in an age when social networks and chat services are absolutely overrun with AI spam, a phone number requirement helps to keep things a bit more under control.

Signal’s design ensures that the phone number requirement isn’t actually the privacy issue people think it is. But more than that, the issues people have with that phone number requirement demonstrates something more disturbing, even depressing: nobody seems to know how to do privacy threat modeling anymore.

§Is threat modeling dead?

To be clear, this isn’t entirely new. For decades, some folks take the all-or-nothing approach: if you’re not completely off the grid, using exclusively open source software, and have a completely anonymous presence on the internet (if you have one at all), then you’re not private enough in their eyes.

The problem with that sentiment is that it scares away the average person. The “normie”, if you will. It’s an unfortunate fact that the majority of people simply don’t care about privacy, and as long as that’s the case, the surveillance state has the numbers and, therefore, the power.

Telling them they have to become anonymous schizos just to become more private is going to rub them the wrong way and tell them that it’s an effort that’s just not right for them. We need people at all levels of concern to join the fight. We need the numbers. Privacy will continue to be a losing battle if we continue to insist on all people going to the maximum.

This is why privacy threat modeling isn’t just important, it’s crucial. Everyone needs to evaluate the aspects of privacy and surveillance that are most important to them. Some people would be content switching to a private browser and using Signal, and that’s fine! They’re already contributing heavily to the cause of privacy, since they’re already doing significantly more than the average citizen is doing.

I’m not saying that people are stupid for going extreme with their privacy. I will say I think a lot of people who do that actually don’t have a legitimate cause to do so, but there are certainly some people who definitely do need an extreme threat model, and that’s also fine. My issue is simply with the fact that there are entire swaths of the privacy community that really do have an all-or-nothing approach because they are genuinely paranoid (even if they won’t admit it).

§Threat modeling isn’t dead

Instead of presenting a problem with only an extreme solution, privacy advocates need to start promoting a spectrum of solutions. We need to start informing individuals of what threat modeling is, and how it helps them ease into the world of privacy without upending their entire life in the process.

“Privacy is a journey, and you have to start somewhere. You don’t have to jump in head-first, you can start with simple things like changing your web browser, using an encrypted chat app, or trying out some open source software.”

That, right there, would catch the ear of the “normies” far more than “The FBI is always listening to you through your phone so throw your phone away, cut off your friends and go live in the woods.”

The truth is that privacy isn’t just a protection against the government. It’s a protection against data theft from corporations. It’s a protection against doxxing by psychotic internet anons. It’s a layered and nuanced problem that requires a layered and nuance response.

§Learning how to threat model

Even experienced users of privacy tools and tactics could benefit from re-learning how to build a threat model.

It really isn’t as challenging or time-consuming as it might seem. It’s as simple as figuring out which specific issues are most worrying to you, then doing research into what measures you can take** to guard against the threat.

If you’re just worried about corporations buying and selling your data, you can take a pretty light-handed approach: switching browsers, using encrypted chat apps, reducing or eliminating big tech services (or isolate them, if you need to use them), switching to a private DNS provider and enabling encrypted DNS. Those things alone, as I stated before, place the average individual at a much more guarded state than the “normie”.

Every contribution is a contribution. Every action has a reaction.**

Of course, if you’re concerned beyond that, you’ll have to evaluate the areas that your perceived threat(s) could hit you the most. For example, if you’re concerned about doxxing, you can eliminate your data from people search sites (or use a service to do it for you), make your social media accounts completely private, wipe the data from archives wherever possible, and so forth. VPNs can transfer trust away from your ISP. Switching to Linux or an Android custom ROM makes OS-level surveillance a non-issue. The list goes on.

The uncomfortable truth for a lot of people using privacy tools is this: many are not just going overkill, they’re actively damaging the cause of privacy and making themselves look far more suspicious** to government agencies in the first place.

§Hiding in plain sight was always the best strategy

Back when Brave switched away from its “strict” fingerprint protection option, I was pretty upset about that at first. However, what I uncovered when I looked it was a very important lesson for privacy strategem in general: hiding in plain sight is a much better strategy than brute-forcing your privacy approach.

In the privacy browser war, I see a lot of people talk about how Librewolf’s fingerprint protections are “better” than Brave’s because it outright blocks canvas fingerprinting. Counter-intuitively, this actually is the wrong way to go about protecting against fingerprinting. Librewolf blocks fingerprinting, but that makes individual users and browser user-agent data stick out like a sore thumb. Meanwhile, Brave’s approach was to create a means of randomizing the browser fingerprint. This means that a site won’t see the same browser fingerprint twice. No red flags are raised, it just thinks two different people visited the site (especially if you use a VPN).

I’m not saying that to dunk on Librewolf. I do use it as one of my browsers; having multiple browsers for different roles is actually the smart play anyways. Instead, I’m using that as an example of the in-plain-sight strategy.

If you just look like an average person, you’re unlikely to be tagged for specialized surveillance. You’ll just look like a random person. This is also the strength of networks like Tor, which are designed to make every user look the same, so that it’s technologically infeasible to pick out individual users without trying other attacks on the areas that the technology can’t protect the individual.

§So, what does that look like for the average user?

It looks like using tools such as Signal for confidential conversations with friends and family that you trust. It looks like sometimes spending in Bitcoin or altcoins, even if you hold your value in Monero. It looks like using Linux as a fun hobby, rather than strictly as a privacy tool. It looks like being careful what you say online when talking about sensitive matters.

People can talk about opsec all they want, but if your opsec causes you to stand out among a crowd (even if you’re anonymized) or if it discourages people from even trying to find a solution for their own concerns, that is extremely counterproductive.

Learn to threat model. It’s good for you, it’s good for the cause, and it eliminates the need to continue chasing the high of the unattainable goal of “perfect privacy”.