The Difference Between "We Don't Log" and "We Can't Log"

“We don’t log your data” shows up in the privacy policy of almost every tool you use. It’s a comforting sentence, and it’s usually true in the narrow sense the company means it — they’re not writing your activity to a spreadsheet and selling it to the highest bidder.

But “we don’t log” and “we can’t log” are two completely different claims, and most companies are only making the first one.

The difference is who’s holding the key

If a company doesn’t log your data, that’s a policy — a decision, backed by internal process and (hopefully) a genuine culture of restraint. Policies are real. They’re also reversible. A new executive, an acquisition, a government subpoena, a well-meaning engineer adding “just one more analytics field” — any of these can quietly turn “we don’t” into “we do,” and you’d have no way to know until it mattered.

If a company can’t log your data, that’s architecture. It means the system was built so that the sensitive information never exists on their servers in a readable form in the first place. There’s no policy to reverse, because there’s nothing to log even if someone wanted to.

What this looks like concretely

When you create a link with PVTLNK, your destination URL is encrypted in your browser with AES-256-GCM before it’s sent anywhere. We receive ciphertext. The decryption key lives only in the URL fragment — the part after # — which is a part of a URL that browsers, by design, never transmit to any server. Not ours, not a CDN’s, not an analytics provider sitting in the middle.

This means if someone compelled us to hand over a specific link’s destination, we’d have nothing usable to hand over. Not “we’d refuse” — we’d be unable to comply even if we wanted to, because the plaintext URL was never something our infrastructure touched.

Compare that to a typical shortener: the destination URL sits in their database in plain text (or with reversible encryption they hold the key to), which means “we don’t look at it” is doing all the privacy work, and that promise is only as durable as the company’s current intentions.

This isn’t a dig at “we don’t log” companies

Plenty of services making that promise mean it sincerely, and a policy commitment is better than none. We’re not claiming everyone else is lying. We’re pointing at a real, meaningful distinction that gets flattened by marketing copy on both sides of the industry — “private,” “secure,” and “zero-knowledge” get used almost interchangeably even though they describe very different guarantees.

If you’re evaluating a tool that claims to protect your privacy, the question worth asking isn’t “do you log this.” It’s “could you log this, even if you wanted to.” The answer tells you whether you’re trusting a promise or trusting math.