How to Protect Yourself from Dark Patterns, According to Dark Patterns Hall of Shame
Updated
Dark Patterns Hall of Shame is a documentation site that catalogs real examples of manipulative interface design, organized by pattern type and company. You can use it defensively: before you sign up, subscribe, or accept a cookie banner, the site's case library shows you what the trick looks like in practice and what the exit path costs. The protection method it supports is not a single checklist but a habit — slow down at the exact moments where interfaces are designed to speed you up.
What the site actually documents
Each entry pairs a named dark pattern with a specific company and a short write-up. The pattern labels recur across cases, which is what makes the site useful as a reference rather than just a complaint log.
| Pattern | What it looks like | Documented example |
|---|---|---|
| Preselection | A box is already checked, opting you into something you didn't choose | Phosus newsletter signup; Groww credit-score check |
| Roach Motel | Easy to get in, hard to get out | LinkedIn Premium cancellation; Amazon Audible cancellation |
| Obstruction | Steps inserted to slow you down | StubHub ticket sales |
| Nagging | Repeated prompts after you've declined | Reddit app download push; TikTok notification requests |
| Trick Questions | Wording that makes you answer the opposite of your intent | Mailchimp sign-up |
| Cookie Wall | Content blocked until you accept cookies | HRAcuity video access |
| Privacy Zuckering | Defaults that share more than you expect | LinkedIn AI training on your content |
| Fake Scarcity | Manufactured urgency | StubHub |
The three moments where protection actually happens
Before you submit a form
Preselection is the most common pattern in the site's catalog, and it's the cheapest to defeat. Scan every checkbox, toggle, and pre-filled field before hitting submit. The Groww case is the clearest illustration: a "pre-approved" loan offer with a "Get started" button, and below it a pre-checked box granting permission to pull your credit score. The button is the decoy; the checkbox is the decision.
The same applies to AI training opt-ins. The Figma entry describes AI features rolled out in early 2025 with all users automatically opted in to have their design files used for training. The LinkedIn entry covers the same mechanism for post content feeding AI models. In both cases the default was "in," and the only protection was finding the setting and reversing it.
When you accept a cookie banner
The HRAcuity case is a cookie wall: a video is inaccessible until you accept cookies, with no equivalent path to decline and still view the content. The practical response is to decide whether you actually need that specific piece of content, and if you do, to treat the acceptance as a one-time, site-specific concession rather than a blanket habit. Accepting on every site you visit is how a single wall becomes a default posture.
When you try to leave
Roach Motel entries are the most expensive to escape because the cost is your time. The Audible cancellation write-up describes a process built from several classic patterns stacked together to complicate the exit. LinkedIn Premium cancellation gets a four-minute read for the same reason. The defensive move is to plan the exit before you enter: check how cancellation works while you're still in the trial or first billing period, not when you've decided you're done.
Why the site is worth revisiting
Dark patterns change as companies change. The catalog spans entries from mid-2024 through late 2025, and the pattern mix shifts — AI training consent appears in the newer entries in a way it didn't earlier. Reading a handful of entries in the pattern category you're most exposed to (subscriptions, if you cancel things often; signups, if you trial a lot of tools) gives you the current shape of the trick rather than a stale list.
The site's own framing is that staying informed is the protection. That's accurate but incomplete: information only helps if it arrives before the click. The realistic use is to treat the catalog as a preview — read the Roach Motel entries before you start a subscription you might want to end, and read the Preselection entries before you fill in a form you'll submit in a hurry.