Ethics

We're a very small company handling something genuinely private: your email. You deserve to know how we operate before you trust us.

So this is the page where we write down our moral framework, what we've decided not to do, and the parts we're still weighing up. No marketing speak, just how we feel.

How we operate

We're self-funded. There are no investors, no board, and nobody pushing us to grow at any cost. People try our service, and if they like it then they pay us. That's the whole business model.

We think that this results in a better product, as it means we only answer to our customers, rather than to someone who wants a return on their money.

We're a small family team. Leave Me Alone is designed and built by Danielle and James, a couple who wanted this to exist and couldn't find it. You can always get through to one of us. We hire freelancers, but we don't outsource the running of the company.

We don't sell your data. Not to advertisers, not to data brokers, not to anyone, at any price. There's no asterisk on this one. Our business is just people paying us for software and always will be.

We self-host what we reasonably can. Everything possible runs on machines we rent and manage. Most companies our size would outsource much more. But everything we bring in-house is one fewer company with whom we have to share data.

We're built on open source, and we say so. Leave Me Alone wouldn't exist without a long list of freely given work, and we credit it on our about page. We choose open tools over closed ones where possible.

How we handle AI

We use some AI, these days there's no escaping it. But we do try to use it responsibly and sparingly.

The majority of our features use no AI at all, but it is used to help train our Spam Blocker, create Rollup and Screener summaries, and help during the unsubscribe process. However, we make the following promises:

Open-weight models, on hardware we rent. The AI that comes close to any personal information is an modèle à poids ouverts Un modèle d'IA dont les poids (le « cerveau » entraîné) sont publiés et téléchargeables par tous. Comme nous l'exécutons sur nos propres machines louées, rien de ce que nous lui envoyons ne quitte notre infrastructure, et aucune entreprise d'IA ne se trouve à l'autre bout pour le collecter ou en tirer des enseignements. , running on machines we pay for. They answer the question and forget it. There's no AI company on the other end of it collecting the data that we send.

Nobody is training AI on your emails. Not Anthropic, not OpenAI, not anyone. When we scan your inbox, summarise a Rollup, or work out whether something is spam, none of it becomes training data for somebody else's model.

We tell you where it goes. Every AI flow in the product, what it sends, and which companies are involved is written out in our privacy policy.

What we wont do

We don't pay for reviews. No money to review sites, comparison sites, or "best of" lists. If you found us on one of those and we came out well, nobody was paid for it. We know how much of that industry works the other way.

We don't use dark patterns. No manufactured scarcity, no hiding the cancel button, and no making you email us to leave.

We do run discounts, and while one is on you'll see a countdown timer on our homepage. It's counting down to a real deadline. When it reaches zero the offer genuinely ends, we don't quietly restart the clock, and you'll never see a banner that's been saying "ends today" since last spring.

Our pricing page explains exactly what you pay and why. If cancelling is ever harder than signing up was, we've done something wrong.

We don't send you email you didn't ask for. It would be a strange thing for us of all people to get wrong. Every email we send has a working unsubscribe link, and it works the first time.

We don't do shady business. No surprise renewals, no burying terms, no quietly changing what we do with your data and hoping nobody reads the diff. When our privacy policy changes we publish the old version and the diff alongside it.

Where we tread carefully

Some things aren't a clean yes or no. These are the ones we've thought hardest about, and we'd rather show our working than pretend the tension isn't there.

Advertising. We're not running ads at the moment and we don't usually. When we do advertise, it's usually through a partner service who shares our ethical framework. For example, see our partners.

We will always take the path of least exposure: no third-party ad or tracking scripts on this website, no pixels following you around, and one first-party cookie that only records which ad you arrived from. If you go on to buy something we tell the ad platform the sale happened, using a empreinte SHA-256 Un brouillage à sens unique d'un texte en une chaîne de caractères de longueur fixe. La même entrée donne toujours le même résultat, mais ce résultat ne peut pas être reconverti en l'original. Cela permet à une plateforme de vérifier si deux personnes correspondent sans qu'aucune des parties ne transmette l'adresse réelle. of your email rather than the address itself. It's all set out under Cookies and advertising in our privacy policy.

Affiliates. Our affiliate scheme isn't running right now, and it has never been open to everyone. We vet who joins, because somebody promoting us with spam or misleading claims does more damage than no promotion at all. The rules are in our terms and we do enforce them.

The companies we do rely on. We can't run this on nothing. We need somewhere to host, a way to take payments, and a way to send email. Every one of those companies is named in our privacy policy along with what they receive, and we keep the list as short as we can manage.

How we grow

By building the product. The money that a funded competitor would spend on acquisition goes into features, infrastructure, and support instead. It's slower. It's also why we're still here after 7 years.

By word of mouth. Most people arrive because somebody told them to. Our Wall of Love is real people who weren't paid or prompted, and we'd rather have ten of those than a thousand bought impressions.

By talking to you. Feature requests, bug reports, and the occasional complaint go straight to the two of us. A lot of what's in the product now exists because somebody emailed and asked.

Holding us accountable

A page like this is easy to write and harder to live up to. If you think we've slipped on any of it, or you've spotted something here that doesn't match what we actually do, tell us and we'll either fix the behaviour or fix the page.

This page was last updated on the 8th August 2026.