When people argue about which tools to use, the security-versus-privacy debate usually goes nowhere because both sides are talking past each other. The arguments for privacy-focused tools (“I do not want Google reading my email”) and the arguments for security-first tools (“I do not want my bank account drained”) are not actually in conflict. They are answering different questions.
This is the article I wish someone had written for me five years ago, when I was bouncing between Signal and Gmail based on whatever a privacy blog had most recently recommended. The mistake I was making, and the mistake most “switch to privacy tools” guides make, is treating security and privacy as the same dimension. They are not.
Why security and privacy are not the same thing
Privacy is about who gets to see your data. Security is about whether someone who is not supposed to see your data can still see it. The two goals overlap when the threat is a large company that you have voluntarily given access, but they diverge sharply when the threat is a stranger.
A tool can be private and insecure. Encrypted email is unreadable to the email provider, which is good for privacy. If the encryption implementation has a bug that lets an attacker recover the plaintext, the email is private from the operator and exposed to the attacker. The user’s threat model determines which failure is worse.
A tool can be secure and not private. A bank uses heavy security (multi-factor authentication, fraud detection, account lockouts) precisely so that strangers cannot get in. The bank also sees every transaction you make, which most people tolerate because the security tradeoff is worth it for the service.
Once you accept that these are two separate axes, the “which one matters more” question stops being a philosophical debate and starts being a threat-model exercise. The answer depends on what you are trying to defend against.
The threat-model framing
Most personal-tool arguments collapse because the participants have different threat models and do not realize it.
- Privacy-first threat model. The threat is the platform itself (Google, Meta, Apple) collecting data that they could be compelled to hand over, sell, or leak. The mitigation is end-to-end encryption and minimization of the surface area.
- Security-first threat model. The threat is an attacker (phishing, credential stuffing, account takeover, malware) breaking into the user’s account. The mitigation is strong authentication, account monitoring, and the use of well-maintained, audited software.
- Mixed threat model. Some users care about both, in roughly equal measure. The mitigation is to pick tools that score reasonably on both axes, even if a specialist tool would score higher on one.
- Sovereignty-first threat model. The threat is a future regime change, a compelled access request, or a vendor deciding to discontinue the service. The mitigation is local-first software (tools that store your data on your own device, not in a vendor’s cloud), open formats, and the ability to migrate without the vendor’s cooperation.
The mistake is choosing a tool based on its privacy score when the actual risk is a phishing attack, or choosing a tool based on its security score when the actual risk is a data broker. Both mistakes are common, and both are visible in the tools people recommend in 2026.
Why privacy-first tools sometimes reduce security
This is the part that “switch to privacy tools” guides skip. Switching from a well-maintained platform to a smaller privacy-focused alternative often means trading security for privacy.
A large platform like Google or Microsoft invests more in security engineering than a small open-source project can afford. The platform has a dedicated security team, a bug bounty program, regular third-party audits, and a security operations center that responds to incidents in real time. A smaller privacy-focused project may have a single maintainer who is also the project lead, the support engineer, and the person who triages bugs.
When you switch from Gmail to a smaller encrypted email provider, you are giving up some amount of security engineering in exchange for better privacy. The question is whether the privacy gain justifies the security loss. For most personal email users, the answer is no, because the realistic threats are credential phishing and account takeover, not the email provider reading their mail.
Why security-first tools sometimes reduce privacy
The other direction is less common but real. A user who values security above everything else will pick the most locked-down option regardless of who runs the infrastructure. A company-managed password manager with strong encryption is more secure than a self-hosted KeePass database, but the company sees (in encrypted form, at least) which services the user has accounts with.
For most users, this trade-off is also not worth it. The realistic threat to a personal password vault is forgetting the master password, not the password manager company being malicious. Choosing a closed-source security-focused product over an open-source alternative means trusting the closed-source vendor not to be the threat you were trying to defend against.
A working priority order
After watching a lot of people make this mistake, including myself, here is the order that actually works for most personal users.
- Identify the realistic threat. Is the bigger risk someone breaking into your accounts, or someone collecting data about you? Pick the threat, then pick the tool.
- Pick tools that are good enough on the other axis. A perfect privacy tool with mediocre security is a worse choice than a good security tool with acceptable privacy, if your threat is account compromise.
- Avoid tools that score badly on both. The worst combination is a privacy-focused tool that is also unmaintained, because it has given up the engineering investment that makes either dimension work.
The first two steps sound obvious. In practice, they require you to be honest about what you are actually trying to defend against, and most people skip that step because it is uncomfortable to admit that the threat they care about is targeted ads rather than a foreign intelligence service.
What I would tell past me
If I could send a message back to the version of me that switched email providers every six months based on whatever privacy blog I had most recently read, three things.
- Pick the threat first, then the tool. “I want privacy” is not a threat model. “I do not want Google to read my email” is.
- A privacy-focused tool with bad security is worse than a security-focused tool with mediocre privacy, unless the threat you actually face is data collection. Most people face credential compromise, not data collection.
- Do not optimize for the worst case. The argument “but what if Google turns evil” is the same argument that justifies any privacy decision, including some that are obviously bad. Optimize for the realistic threat.
Trade-offs
Prioritizing security first is not a free choice. It means accepting that the platform you use sees some of your data in exchange for the engineering investment they make in keeping attackers out.
In our case, the trade-off worked because the realistic threats I face are credential phishing, account takeover, and the occasional malware, not platform-level data collection. Your math will be different if your threat model is “I do not want any single entity to know my reading habits” or “I am in a profession where targeted surveillance is a realistic risk.” The shift in mindset took about a year, with most of that going to actually writing down what threats I was defending against before picking a tool. One specific thing that was harder than expected was admitting that the platform I was using was already reasonably secure, and that switching for privacy reasons was trading a real benefit for a marginal one. One thing that was easier than expected was discovering that a clear threat model makes tool decisions dramatically faster.
If your realistic threat is credential compromise and account takeover, security-first is the right priority. If your threat is data collection by a specific platform you have chosen not to trust, privacy-first is the right priority. If you do not know what your threat is, write it down before you pick a tool. The act of writing it down is the part that fixes most bad tool decisions.