BERLIN, 24 March 2021 — Xayn, the privacy-protecting search app that utilizes AI to deliver personalized and relevant results, has been downloaded more than 100,000 times since its launch in December. In a time of increased online activity and with debates about online privacy heating up, Xayn has found a winning formula for web search: its sleek interface and user-trained AI delivers accurate results, while always protecting personal data. Since its launch three months ago, Xayn’s daily search numbers have risen to more than 90,000 – a number that took DuckDuckGo three years to achieve. The success of this academically grounded startup ushers in a new generation in privacy tech.
The European tech company combines privacy and a convenient search with control over the search algorithms, empowering users to take back control of their data. For the first time, you can influence what is shown to you by swiping left or right on the results. This way, you can teach the algorithms which results you want to see more of in the future.
Growing pressure from consumers for more privacy-protection
The growth of Xayn comes at a time when internet users are becoming increasingly aware of the intrusive and exploitative nature of Big Tech. These companies have in turn responded with heartfelt statements and pledges that sound nice but don’t always lead to action. There appears to be a growing consensus that personal data should be controlled by the person generating it. A new model is emerging where online access does not entail surrender of digital privacy.
Current data such as the Cyber Security Insights Report demonstrates this trend towards more privacy: 92% of globally surveyed online users stated that they have at least one significant concern about their data privacy. More than two-thirds of those surveyed (67%) are more alarmed than ever about threats to their privacy but 64% still accept certain risks to their online privacy to make their life more convenient. Other numbers indicate the relevance of search for how people use the internet: Search queries make up around 30% of all global website traffic.
From research to search
"The first search engines were developed by scientists in order to make online information available to everyone," highlights Leif-Nissen Lundbæk, co-founder and CEO. "Over the last 30 years this system has unfortunately been perverted, forcing users to become losers – you either lose your online data or your precious time when you don’t find what you’re looking for. I was fed up with always having to choose between privacy and convenience when searching online. To end this trade-off, we built a search alternative with no compromises. With Xayn, you’ll find the information you’re looking for in an easy way while always keeping your privacy and deciding yourself – that’s web search as it should be.”
Born in academia, the company behind the new search engine was founded to move an entire industry forward. To this day, that academic vision remains. Xayn currently plays host to a workforce of 30% PhDs who are committed to privacy as a basic human right, as stated by the United Nations in their Universal Declaration of Human Rights.
"Following in the footsteps of the scientific web pioneers, we not only wanted to make information easily accessible but also privacy available to everyone. That’s why we pioneer a new generation of user-friendly privacy tech made in Europe,” explains Michael Huth, Professor at Imperial College London and co-founder and Chief Research Officer. “We have a strong focus on design because we want to offer users the best experience possible. We are convinced that privacy tech can only go mainstream if users are given alternatives that can compete with the incumbents on design and user experience.”
Four AI models working on your device to find the best results
Unlike its competitors, Xayn’s web search uses a decentralized approach and trains its artificial intelligence directly on your devices. Instead of bringing the data to the algorithm, it brings the algorithms the devices. All personal data stays where it belongs – with you. Only the AI models are shared after they’ve been encrypted to enable collective learning. This decentralized approach is called masked federated learning. It is made possible by the company’s platform XayNet which started as a research project four years ago and which has been open-sourced since then.
Every time you search online with Xayn, four different AI models work on device level together to personalize your search results. The first model processes natural language to understand each search’s context. The second model creates dynamic clusters of interests based on your input. The third model then analyses the domains you like or don’t like. The fourth model mixes all this together and adds more contexts like the time, weather, location, and other variables to find the best matching search results. This last model also acts as a protection against filter bubbles because it will explore new content outside of your past preferences and show it to you.
By swiping left or right you can give the AI direct feedback and train it to give you even better matching results in the future. The more you use Xayn, the more accurate and personalized the results become. All of this is done, again, without collecting any of your data.
With Xayn being all about privacy and innovation through technological achievement, the company is continuously looking for ways to ensure the best user-experience in web search. Therefore, Xayn offers you features such as:
Xayn search app is available on all smartphones (iOS, Android) and will also be available in a web version at a later stage. Users can download Xayn free of charge from Apple’s App Store and the Google Play Store.
Xayn is a privacy-protecting search engine that enables users to gain back control over the algorithms and provides them with a smooth user-experience. Using the latest AI technology made in Europe, the company ushers in a new generation of user-friendly privacy tech – making privacy available for everyone.
The AI company started as a research project at The University of Oxford and Imperial College London by Leif-Nissen Lundbæk (Ph.D.) and Professor Michael Huth. Together with Felix Hahmann, they founded the tech company in 2017. To this day, that academic vision remains with a workforce comprised of 30% Ph.D.s. The company’s open-source framework for federated analytics and learning, XayNet, is the basis of the privacy-protecting personalised search alternative Xayn. The Berlin-based company has received investment funding of 9.5 million EURO by Earlybird VC as well as Dominik Schiener. Xayn has worked with corporations such as Porsche, Daimler, Deutsche Bahn, and Siemens.
TECHNOLOGY AND FUNCTIONALITY
How does Xayn work?
Xayn uses a decentralized approach with so-called computing on the edge. The app is based on XayNet, a specially developed framework for masked federated machine learning. The app doesn’t collect any user data but combines the encrypted AI models of the respective devices asynchronously with homomorphic encryption to a collective model. In a second step, this collective model is then fed back to the individual devices. Thus, the respective models learn from each other while your privacy is always protected.
On each device, four AI models work together in the app to present you the most relevant search results.
In the beginning, Xayn delivers general search results but refines them every time you interact with it. This way, you can watch the AI learning and get relevant search results, while your privacy is always protected.
Why should you interact with the Xayn app by swiping?
With each search, you can give Xayn direct feedback so that the app refines future search results. If you swipe to the left, you’re telling Xayn you don’t want to see the result. With a quick swipe to the right, you tell Xayn that you're interested in the result. Swiping slowly to the right indicates to Xayn that you want to save the result in your collections.
What are the Collections?
When searching the web with Xayn you can create flexible content collections where you can save webpages, news, images, videos etc. These personal content collections provide you with all their researched information at one glance, in an interface that provides a clear overview.
How does the personalized home screen work?
Like a social media newsfeed, your personalized home screen offers you a feed of curated content that matches your interests – be it general web content, videos, images, or news articles. However, unlike conventional new feeds, Xayn doesn't rely on a closed platform but rather collects content suggestions uniquely for you from the variety of the entire web. With every app interaction you make, Xayn understands better what content is relevant for you.
What are the requirements for the Xayn app on a smartphone?
We recommend you use the Xayn app with a smartphone that is as up to date as possible - for Apple starting from the iPhone 8 and for Android operating systems from version 9.
Is Xayn also available as desktop version?
A desktop version of Xayn as well as other features are already in preparation and will follow in the future.