At the bottom of AlignAlgo's manifesto page sit five short lines, easy to skim past: ads only on your terms. No pay-to-rank. See every parameter. Export and leave anytime. AI always labeled. They read like a values slide — the kind of list a lot of companies put up and then quietly walk back once growth gets hard. We'd rather explain what each one actually commits us to, in plain terms, so it's a promise you can hold us to instead of a mood board.
"Ads only on your terms"
Updated August 12, 2026 — this promise started life as "no ads, ever." We've sharpened it, and a transparency movement says so out loud instead of quietly editing the page. The absolutist version took a choice away from you: some people genuinely want to hear about products that fit their life. The real promise was never "no commerce" — it was "no one buys your attention behind your back."
Most consumer platforms run on a simple, mostly invisible trade: your attention is the product, and advertisers are the actual customer. That's not a conspiracy — it's just the business model, and it explains almost everything about how those apps are designed. Feeds optimized for time-on-app. Notifications tuned for the moment you're most likely to come back. Recommendations that quietly favor whoever's paying for the placement over whoever's actually best for you.
"Ads only on your terms" means we're not building toward that trade — the commercial layer has a dial, and you're the only one holding it. Block it all and browse with zero commercial suggestions, forever; that mode isn't going anywhere. Or open the door on your rules: Inform Me, Best Value only, Ethical only. Every commercial suggestion is labeled as one, can explain why it showed up, and can be dismissed or tuned. What never changes: no advertiser sits between you and your recommendations, and nobody can pay for the slot.
"No pay-to-rank"
This is the sharper, more specific version of the ads promise, and it's worth separating out because it's the one that's easiest to quietly cheat on. Plenty of platforms that don't run traditional "ads" still let sellers, restaurants, or services pay to move up a ranking — a sponsored listing dressed up to look like an organic result. It's a subtle dishonesty: the ranking still looks like it's about quality, but part of it is actually about who wrote a check.
No pay-to-rank means the order you see things in reflects what the system actually believes fits you — total cost, real quality, trust signals, fit for the moment — never a bid. Nobody can buy their way up the list. If something ranks first, it's because the system thinks it should, full stop.
A recommendation that can't explain itself is a black box wearing a percentage.
"See every parameter"
This is the transparency half of the promise, and the one most directly tied to the Own Your Algorithm pledge. A "92% match" that shrugs when you ask why isn't actually information — it's just a number dressed up to look authoritative.
"See every parameter" means the reasoning behind a ranking or a match isn't hidden by design. You should be able to ask why something was recommended and get an actual answer — which factors mattered, and roughly how much — instead of being told to trust the number. That's a genuinely different engineering commitment than most recommendation systems make, because most of them are built assuming nobody will ask, and this one is built assuming you will.
"Export and leave anytime"
Data that can't leave with you isn't really yours — it's on loan, revocable the moment you stop being useful to whoever's holding it. A lot of platforms make this technically possible but practically miserable: a data export exists somewhere in settings, buried three menus deep, arriving as an unreadable file nobody actually uses, clearly built to satisfy a regulation rather than to be genuinely usable.
"Export and leave anytime" is the opposite intention: the graph of what you've liked, trusted, and chosen is portable because it was designed to be portable from day one, not retrofitted under pressure. If you decide to stop using AlignAlgo tomorrow, what the system learned about you doesn't just evaporate — and it doesn't hold you hostage either.
"AI always labeled"
AI-generated content and AI-assisted recommendations are quietly everywhere now, and increasingly indistinguishable from the human or purely-algorithmic version, on purpose, because ambiguity performs better in engagement metrics than honesty does. We think that trade is exactly backwards. If an image is AI-generated, it says so. If a recommendation involved an AI system making a judgment call, that's disclosed too. Not because regulation requires it everywhere yet — in most places, it doesn't, not fully — but because the whole premise of Own Your Algorithm collapses the moment we start hiding what the algorithm is doing, even in small ways.
Why this is the actual product, not a values footnote
It would be easy to read these five lines as marketing — the kind of promise that's cheap to make and expensive to keep, which is usually a sign it won't survive contact with a hard quarter. We'd rather you treat them as a specification. Every one of them is a constraint on how AlignAlgo can be built, not just how it's described: no advertiser relationship to protect, no sponsored-slot revenue to defend, an explainability requirement baked into the ranking logic itself, an export path that has to actually work, and a labeling discipline that doesn't get quietly dropped once nobody's watching closely.
Five short lines. The whole rest of AlignAlgo — the graph, the family of tools, the pledge — is what it looks like to actually build a product that keeps them.