Full transparency: Links marked with (*) are affiliate links. Yes, we might earn a commission if you buy. No, that doesn’t mean we’re shilling garbage. We recommend what we’d actually use ourselves. Read our full policy.

The premise

S/MIME is an open standard. It shouldn’t cost anything beyond the certificate itself, and a handful of CAs — Actalis, HARICA — offer free tiers for individuals. On paper, “secure, signed, encrypted email for anyone” looks achievable without an enterprise budget.

What actually holds up

  • The standard is sound. S/MIME certs carry the emailProtection EKU, are validated through mailbox-control challenges (click a link, prove you own the address), and work natively in Outlook, Thunderbird, Apple Mail, Gmail — no plugins.
  • Free issuance is real but narrow. Actalis grants one free Mailbox-Validated cert per account. HARICA’s free tier is scoped to academic institutions. Neither is a sustainable path for multiple identities or a team.
  • Self-hosted CAs solve the cost problem but not the trust problem. A private root (step-ca, easy-rsa) issues valid certs for free, but only clients that have imported your root CA will trust them. Fine for a closed circle of known contacts. Useless for “anyone.”

Where it breaks in practice

Two structural issues, not edge cases:

  1. One cert per identity. Multiple people sharing one certificate means sharing one private key — anyone with it can decrypt anyone else’s mail and impersonate the sender. Every distinct identity (family member, team account, alias) needs its own cert. That’s where “free” quietly becomes “per-seat.”
  2. Corporate mail gateways strip it. Enterprise antivirus/DLP scanners decrypt, inspect, and re-wrap mail at the perimeter. The original signature no longer validates on arrival. Send a signed/encrypted email to any organization running standard perimeter security, and the recipient sees a broken seal — through no fault of your setup. This is a known, unresolved friction point in S/MIME adoption generally, not something specific to a DIY approach.

Why Let’s Encrypt’s model doesn’t transfer

It’s tempting to ask why LE can’t just do this for free the way it does TLS. The honest answer: it could, technically — MV-tier S/MIME validation is just as automatable as LE’s domain-control challenges. What’s missing is the funding model. LE’s sponsor coalition (Mozilla, Google, Cisco, AWS, Meta) subsidizes it because a fully-encrypted web serves their business interests directly. No equivalent coalition exists for personal email encryption — so the automation exists, but nobody’s paying for it at scale, and commercial CAs treat their free tier as a lead-gen loss-leader rather than public infrastructure.

Conclusion

S/MIME remains worth deploying for narrow, high-value cases — legal, medical, financial correspondence with a known, closed set of recipients who can import a shared root, or where the few euros per identity for a commercial cert is trivial against the value of non-repudiation. It is not a general-purpose “secure email for everyone” solution today: the per-identity cost model and the enterprise-gateway problem both work against it at any real scale.

For routine correspondence, standard TLS-in-transit (STARTTLS) is the pragmatic baseline. For anything that genuinely needs strong guarantees, an out-of-band channel (Signal, encrypted file transfer) sidesteps the gateway problem entirely.