Alex Polyakov is the CEO and founder of Adversa AI. He has spent 18 years in deep technology as a cybersecurity researcher, well-known hacker, serial entrepreneur, and trusted AI evangelist. It is a renowned Israeli business that develops applied AI security techniques. Its goal is to increase public confidence in AI while defending it against safety hazards, privacy concerns, and cyberattacks.
Alex went over every position in cybersecurity, starting with a researcher who disclosed 300+ zero-day vulnerabilities in several platforms even before bug bounties were introduced. He delivered some of the most intriguing research and thrilling hacks as a pen tester at security conferences across the world, such as BlackHat, RSA, and HITB. Once he realized that hacking was simple but defending was difficult, he improved his talents in other areas and experimented with various roles. Alex rose through the ranks to become a product manager, CTO of an AppSec firm, conference organizer, and ultimately CEO of his own startup.
Below are highlights of the interview:
Tell us about yourself and your story before starting Adversa AI.
I’m a son of an astrophysicist engineer and a non-traditional healer. I suppose this kind of duality played a vital role in my curiosity and willingness to understand the entire world. When I was a kid, being surrounded by books about Osho meditation on one side and black holes on the other, I was always curious about how things worked. My interest in math and computers quickened thanks to my father. Furthermore, noteworthy is the fact that my aunt was an Orthodox Jew, instilling Jewish traditions in me. Thus, I realized how many contradictory concepts existed in one civilization, or at least in one family.
I always had an entrepreneurial mindset and earned my first money when I was nearly 7 years old and made candles from old wax to sell them on the market. Therefore, when I chose a specialty in the university, cybersecurity was the only option that suited me. It was the best decision. I found a place where I could constantly look at new systems, learn how they work, and then try to figure out how to make a system work in an unexpected way.
Why did you become a cybersecurity researcher to work in?
Cybersecurity is the most engaging and challenging role in IT. A researcher’s goal is to explore complex systems and understand how they work. Furthermore, you are usually limited in terms of available information and time.This brings the need to apply experience and creativity and invent new methods. I remember people at the university complaining about seemingly useless esoteric programming languages, but I was eager to use them. It was another penny piece of the puzzle in the whole range of possible types of programming languages.
Tell us more about Adversa AI.
When I figured out the next move, my team and I already tried multiple ways to apply AI to various cybersecurity tasks, and we saw that AI itself could be hacked. There are no ways to deal with it once you realize that the critical invention that is exponentially invading the world and taking responsibility for our lives is vulnerable. Hence, we started Adversa with a mission to use everything we learned about cybersecurity to help AI users and developers build more robust and secure AI.
Adversa AI is an innovative Israeli AI startup striving to increase trust and security in AI. With its patented technology, Adversa AI provides solutions for mission-critical AI-driven industries, including but not limited to the Internet, Finance, Smart Cities, and Biometrics.
What are your responsibilities as the CEO of the company?
As a typical CEO, my main responsibility is to support the company’s operations and identify opportunities for the company to achieve its goals. If formally, my responsibilities can be divided into three domains: 1) Leadership – everything related to setting a vision, long-term objectives, and guiding the mission; 2) People – relationships with the team, partners, advisors, peers; 3) Resources – finding resources for the company to grow and continue innovations.
What have you failed at, and how do you overcome challenges?
I constantly fail. It happens every day. If not, it means that I make progress slowly or that I don’t challenge myself to bring our business to success. The question of overcoming challenges is vital. I was “lucky” enough to have a lot of challenging experience. I traveled to 60 countries, nearly died three times, and tried most extreme sports, breaking many parts of my body.
I’m not sure that I can say that I know how to overcome challenges. However, psychotherapy, meditation, and mind-altering exercises are definitely essential if you are in a chaotic and stressed environment.
What are the critical attributes of becoming a successful cybersecurity leader?
The critical attributes of becoming a successful cybersecurity leader are the same as for any other leader – balance is the key. A risk management mindset helps to focus on prioritization and control threats. A technical background is required to earn the trust of a great technical team and motivate, Communication skills open the door to new opportunities. These three traits can lead to success.
What changes do you wish to bring to society?
I love saying that I break artificial intelligence in the daytime and cure natural intelligence by night. Apart from the things we do at Adversa AI on increasing trust in AI, I think our society needs to make more effort and help people overcome psychological traumas. An enormous number of people live with depression and PTSD, which is unacceptable today. I wish it could be changed.
In your opinion, what are the primary technologies that have the potential to revolutionize the future of the cybersecurity sector?
Artificial intelligence might help, as there won’t be solutions without AI. But only if you build it in the right way. There are many issues with AI that can lead to catastrophic consequences. AI is vulnerable to manipulation attacks, so you can fool the decision process. It is prone to infection attacks, so you can modify it and control its actions. Extraction attacks on AI are also a concern. You can steal data or secret algorithms that AI uses. Moreover, you can fool computer vision based on AI so that it fails to recognize a stop sign. The same thing you can use to bypass an AI-driven cybersecurity tool.
What advice would you give to the next generation of security leaders?
Be curious and look at other areas for inspiration: how nature lives, how the brain works, and how the immune system deals with viruses. This is a promising area from the technical perspective. Undoubtedly, take care of your physical and mental health.
What are your plans to sustain the company’s success?
AI is our future, but we can’t trust it nowadays. That’s why the world may experience a new AI winter. I know security, and my contribution is to make AI more robust. I am also sure that it will become more accurate if we make it secure.
AGI is expected to be here. If it goes out of control, we will be able to connect or hack it to save humanity.
Website: www.adversa.ai
