Our story

We built the ad blocker
AI platforms didn't want you to have.

AI platforms started selling ads inside their answers. Not beside them. Inside them. We think that's a problem worth fixing.

47k
Active users
10
AI platforms covered
2,400
Filter rules
MIT
Open source licence
The problem

AI platforms discovered they could sell ads inside the answer - and no one was stopping them.

When you ask a traditional search engine a question, the ads are easy to spot. They sit in a labelled box above the results. You've learned to skip them. The industry calls this "banner blindness" - and advertisers hate it.

AI platforms solved this problem by eliminating the distinction entirely. When ChatGPT recommends a product, when Perplexity cites a source, when Gemini suggests a service - there is no labelled box. The recommendation is the answer. The ad is the response.

The disclosure - when it exists at all - is a 10px grey label in a corner. Below WCAG contrast minimums. Designed to be missed.

This is a fundamentally different problem from web advertising. Tools like uBlock Origin are extraordinary at what they do, but they were built for a different web. CSS class names from 2015. Network requests from known ad servers. Standard banner dimensions.

AI ad formats don't look like that. They're conversational. They're dynamic. They're served from the same endpoints as legitimate content. They need a different kind of blocker.

How it started

It started with one Perplexity answer that felt wrong.

In mid-2024, one of our team members noticed something unusual in a Perplexity response. They'd asked about noise-cancelling headphones. The top result - presented as an AI recommendation - was for a product the model had been paid to suggest. The disclosure was there, technically. But it was designed to be invisible.

We opened the browser's developer tools. The sponsored container had a data attribute. We wrote a quick CSS rule to hide it. It worked. We showed it to a few people. They asked us to make it easier to install.

That was the first AdNeutral filter rule. There are now over 2,400.

We published the extension on GitHub in August 2024 with a single README and a handful of rules for Perplexity and ChatGPT. Within two weeks it had 3,000 stars. Within two months, 47,000 installs. We hadn't done any marketing. People were just sharing it because the problem it solved was real.

We introduced Pro subscriptions in early 2025 to fund full-time development. The free plan - covering the 5 biggest platforms - will never change. We made that commitment publicly on launch day and we intend to keep it.

What we stand for

Five things we won't compromise on.

Free blocking stays free.

The free plan has covered the top 5 AI platforms since day one. It always will. We don't cripple it to drive upgrades.

We don't sell your data.

We don't collect browsing data, AI conversation content, or filter match logs. The stats dashboard is local-only. Our business model is subscriptions, not surveillance.

Everything is auditable.

The extension code, the filter rules, and the rule update pipeline are all public on GitHub under MIT. You don't have to trust us - you can read what we ship.

Zero performance impact.

Filter rules are compiled declaratively. No content scripts running in every tab. No background polling. AdNeutral adds <1ms to page load time on any device.

Rules stay current.

When an AI platform changes its ad format - and they do, regularly - we update the rules. Pro users get updates within an hour. Free users get weekly rule syncs.

Community-first development.

Our issue tracker is public. Our roadmap is public. Every filter rule change goes through a pull request. The community finds new ad formats before we do - and we welcome every report and contribution.

Open source

Read every line. Fork it. Improve it.

AdNeutral is MIT-licensed and fully open source. The extension, the filter rule lists, and the update pipeline are all on GitHub. No "source available" gimmicks. No private core. The code you install is the code you can read.

We accept pull requests for new filter rules, browser compatibility fixes, and performance improvements. Our contributing guide explains the rule syntax and testing process.

The team

Small team. Big filter list.

AdNeutral is built by four engineers who got frustrated with AI ads and decided to do something about it.

JK
Jamie K.
Extension Engineering
SR
Sara R.
Filter Rules & Research
TN
Taiki N.
Backend & API
PV
Priya V.
Design & Web

Plus 38 open-source contributors who have submitted filter rules, bug fixes, and browser compatibility patches.

Press & media

As covered by

"AdNeutral fills the gap that traditional blockers like uBlock Origin miss: ads embedded directly inside AI responses."

- The Verge

"3,000 GitHub stars in two weeks. The frustration with AI ads is clearly real and widespread."

- Hacker News #1

"A technically credible extension from a team that understands how the new wave of AI advertising actually works."

- Ars Technica

Press enquiries: press@adneutral.com  ·  Download press kit

Get in touch

We actually reply.

Found an AI ad format we don't block yet? Want to report a false positive? Have a question about Team pricing? Send us an email.

Ready to start

Your AI answers shouldn't be for sale.