SafetyDetectives spoke with Michael Levit, CEO and founder of Tempest, about data privacy in search, how AI is impacting the search industry, why the Founder’s Pledge is important to him, and much more.
If we could get started by hearing some background about yourself and what motivated you to start Tempest?
I entered the world of ad tech and search engine marketing around 20 years ago. It began with a deal I facilitated between AOL and Google, which happened almost by accident. Back then, AOL had a massive audience that used it for instant messaging, and when they would install the AIM software, we prompted them to install the Google toolbar.
Interestingly, people tend to take the easiest path to get what they want without reading the fine print. In this case, they simply wanted the new version, and it changed their search settings At that time, search was considered a commodity, and people would accept whatever search engine their browser had. This is still largely true today. People weren’t really considering the implications, and Google took off. That deal played a pivotal role in their success.
I worked on various projects involving browser add-ons, toolbars, and extensions. However, I realized that people weren’t paying enough attention to their data and its handling, which bothered me. My previous company generated significant revenue and directed substantial traffic to Yahoo, where users were monetized well. However, I didn’t believe they were being treated fairly.
So, at the core, I founded Tempest with the ideology of creating a search experience that respects the consumer. We still want to give them a world-class search experience, so no one has to choose between having a great experience or maintaining their privacy—a choice we’ve mostly been forced to make until now. This intention has driven us since day one, and it’s incredibly exciting after four plus years, we’re finally here.
What’s the advantage of using Tempest instead of something like Google or Yahoo?
Let’s talk about Google, which is the 1000-pound gorilla in the search market with almost 90% of the market share. There are about 10 significant ways in which they collect data, with one of the major methods being the collection of everything you search for.
Additionally, nearly 90% of users utilize Google’s browser, Chrome, which collects every click and tracks your online activities. This vast amount of data is used as a signal to determine the most relevant ads for you. While this may sound great in terms of personalized ads, it raises concerns about user choice and privacy. 10 years ago, such practices would have been called spyware – today we call it ad tech. I don’t think it’s fair to users that they are not given a choice regarding the information being collected about them, where it goes, and how it’s used.
Many people have this feeling that they’re being watched, and the reality is that they are. For example, if you perform a search on Google and then visit an unrelated website the next day, you may encounter an ad related to the topic you searched for previously. It creates a sense of being watched or retargeted, which can be uncomfortable.
To us, this resembles Big Brother, it’s not a pleasant experience, and therefore Tempest does not do this. However, we don’t think advertising is a bad thing and we believe in ad tech, because a search engine needs to earn money to operate as a business, and I don’t think users really want to pay in order to have a search engine. We use a method which we call ephemeral search, which means that you still have advertising, but it won’t follow you around the web. The ads are only based on the keywords you’re searching for, and once you leave Tempest, for the most part, we forget about you.
I say “for the most part” because we do retain a minimal amount of search data, not related to ad tech, primarily to confirm that there are no bots involved. However, we don’t use this data for future targeting in any way, shape, or form.
Regarding security and privacy, what are the biggest risks people take when using a standard search engine or browser?
One of the biggest risks is having everything you do recorded and stored in a giant database. I think that one of the biggest things that no one really talks about is what would happen if the Google database is hacked. Although the reality is that the risk of that’s pretty darn small since they are one of the more sophisticated companies in the world, and they encrypt your data. So this doesn’t worry me too much
However, what I am more worried about is what happens to the data after you make a search. For example, if you do a Google search about yourself, it gets passed from you to your ISP, and it can be exposed during the transmission, and used in retargeting.
Imagine searching for medical information related to a health condition you’re experiencing. Many people turn to search engines to diagnose their symptoms. Advertisers and other entities can bid on and gain access to this sensitive information to varying degrees.
There’s also a concept called browser fingerprinting, which occurs even if you’re browsing in incognito mode, which gives a false sense of security. This is because various aspects of your system configuration, such as operating system, language, patches, font size, keyboard defaults, and more, contribute to a unique fingerprint. By combining these factors, advertisers can track and target you effectively. It means that even though your browser attempts to hide your identity, the data you unknowingly provide makes it possible to trace and target you for ad filtering and personalization purposes. This can narrow down your exposure to certain news stories or affect pricing on e-commerce platforms like Amazon, based on what they know about you.
Numerous companies are constantly seeking ways to extract more value, be it an extra 10 cents or a dollar, by leveraging user data. However, from a fairness perspective, it raises questions about healthcare privacy, pricing manipulation, and the sharing of personal information. These practices rely on different levels of targeting, whether it’s through fingerprinting or overall website tracking.
How is the Tempest browser different?
We started by saying that there is nothing wrong with Google Chrome as a browser, but it has a whole bunch of bad stuff in it. I like to use lasagna as an analogy, which has many different layers. Chrome has all these excellent layers, like fast rendering, great page layouts, but at the top is a terrible layer that tracks everything you do.
What’s great about Chrome is that it’s available as Chromium open-source, Google gives away all the source code. So we, at Tempest, took the code, ripped out the top layer of tracking, which isn’t a nightmare as long as you know what you’re doing.
The challenge lies in keeping up with Google’s frequent updates and security patches. Every two weeks, they release a new version of Chrome with various security updates. If it takes us two weeks to remove the tracking layer and a new Chrome version is released, we would fall behind, potentially leaving security vulnerabilities unaddressed.
To overcome this, we have developed proprietary technology that leverages artificial intelligence. This allows us to analyze the differences between the previous version and the latest Chrome version quickly. With this automated process, we can release a Tempest browser version within hours instead of days. We’re able to take out that kind of spyware tracking layer at the same time get all those wonderful goodness layers of lasagna that are often tasty.
How do you see the future of the internet and search evolving in the next couple of years?
Search has been relatively stable for almost 20 years now. Innovation in search has mainly focused on verticalization, where search experiences are tailored to specific domains such as travel, news sports etc. However, in the last six months, that’s all changed. We’re seeing the rise of artificial intelligence, the one that’s most often cited is GPT. This is also the most exciting and also the most terrifying thing I’ve ever seen.
What we’re starting to see is the intersection of search and AI, but the AI still isn’t really good enough. For example, let’s say you ask it “what’s the best restaurant in New York City right now?” It will give an answer, and you might agree or disagree, but there is no data source to back it up. Now, if you ask the same question to a search engine, it will give you a list of 10 restaurants with supporting data, such as reviews, star ratings, and pictures.
As the AI improves, it will start to learn your tastes individually. At some level, this is dreamy, it understands me, it knows these things. But now there’s a hell of a lot of personal data that’s going to go into search in a way that can’t really fathom, and we don’t know how the data will be treated, as the AI interacts with thousands of content partners and billions of data inputs.
Ultimately, I think we’re going to need very ethical companies with very clear statements of how your data will and won’t be used. By the way, at Tempest we will implement AI but we will do it extraordinarily slowly, because we want to understand everything we’re doing rather than throw caution to the wind. So, perhaps we’ll be slower on some search results, but at the same time, we’re going to protect users because I think some of the things being pushed out today are reckless.
Another big trend in search is the combination of search and AI with the integration of voice assistants. Companies like Amazon, Apple, and Google are bringing together search, AI, and voice assistants, resulting in a comprehensive user experience. This convergence of GPT/AI, search, and voice assistance is a remarkable development that will shape the future of Internet search.
I noticed that you are participating in the Founder’s Pledge, can you talk about what that is, and why it’s important to you?
My last company was really focused on making money, and thankfully I was very successful at that thing. I continued onward from there and said with this new company, Tempest, I really want to be about making an impact on the world.
I still care about making money, but the reason I care about making money is not for myself. It’s because I have a team of people, and I believe in capitalism, I believe that they’re going to work harder when they get money to take care of them for the rest of their life. So, my hope is that one day we can take Tempest public, make a lot of money, and have it owned by the public. This would be aligned with my personal ethos of looking at the long term good, instead of the short term profit.
And then for my personal stake in it, I want to have some play money, so I’m not going to give the whole company away, but I decided to give away the majority of the company by putting it into a charitable foundation.
The Founder’s Pledge is something where you agree that you will give a certain percentage of the company to charitable endeavors. For me, I’m terrified of climate change, so I will be giving away roughly half of Tempest into climate endeavors over time. I don’t know what the timing is yet, but I’m guessing it will be in around 10 years, and then I will hand over the baton and have someone else run Tempest. I hope to work on some kind of philanthropic charity, where I will take this trust and invest it in the climate.