"Dual-stack" describes the actual state of most modern internet connections: both IPv4 and IPv6 running side by side, available at the same time, with your device deciding per-connection which one to actually use.
Why dual-stack exists
Rather than a hard cutover from IPv4 to IPv6 — which would require every piece of internet infrastructure to switch simultaneously, an impossible coordination problem — the industry adopted a gradual coexistence model. Networks and devices support both protocols, letting the transition happen incrementally over years without breaking anything that still depends on IPv4.
Address selection
When a dual-stack device makes a connection, it typically follows a preference order defined by its operating system — many modern systems prefer IPv6 when both a IPv6 and IPv4 path are available and working, falling back to IPv4 automatically if IPv6 isn't reachable for a specific destination.
Fallback behavior
This automatic fallback is generally seamless from a user's perspective — you're rarely aware which protocol a given connection actually used. It's also exactly why the two paths can behave differently without being obvious: a VPN protecting IPv4 while leaving IPv6 exposed produces no visible symptom, because the fallback logic just quietly uses whichever protocol responds.
Why the two paths can differ
Because IPv4 and IPv6 are functionally separate networking stacks, they can have entirely different routing, different VPN tunnel coverage, and different DNS behavior — treating them as one unified "internet connection" obscures this, which is exactly why dedicated dual-stack testing (see Checking IPv4 and IPv6 Consistency) matters for anything privacy-sensitive.
FAQ
Is dual-stack a temporary arrangement?
It's been the practical norm for years already and is expected to remain so for the foreseeable future — not a short-term transitional state about to resolve into IPv6-only.