Q&A with Ben Colman, Co-Founder and CEO of Reality Defender

Ben Colman is the Co-Founder and CEO of Reality Defender, a leading provider of deepfake detection. Ben worked at Goldman Sachs from 2015 to 2017 in Engineering in New York.
 

1. How did you start your career at Goldman Sachs and what was your path at the firm? 
I was finishing grad school at NYU and splitting my time between New York and San Francisco, when I co-created the first blockchain-based voting platform, PlaceAVote. This project allowed me to join some interesting meetings with government, defense, and intelligence, as well as industry players. I attended a lunch hosted by GS and led by a former GS CTO who was visiting San Francisco to meet data-driven startup founders. I had not previously considered a role at GS but learned that the firm was more of a tech company than anyone realized, and certainly had more of an entrepreneurial spirit in tech when compared to any other bank. I was intrigued, to say the least. 

I then had the chance to visit the GS office when I was in NY where I had a few informational interviews, which led to several great conversations about the firm’s engineering philosophy and my own career plans, but there was no job that matched my experience and interests. I continued informational interviews until I was notified about a role within the Office of the CIO that would be cross functional between technology commercialization, strategic investment, and any and every daily urgency that could (and would) come up with the CIO, CTO and CISO, especially as Engineering would collaborate with client-facing groups across Investment Banking, Private Wealth Management, and the then Securities Division.

I took this role and quickly began to see the incredible “One GS” mentality of how completely different divisions (Banking vs. Engineering) collaborate with multiple dotted lines constantly growing to support clients.

2. What did you learn during your time here that helped you in the next phase(s) of your career?
My time at GS was incredible. I sat less than ten feet from the offices of the CIO and CTO and directly in front of Phil Venables, the CISO at the time, who would soon expand my vision on the opportunities across the power of cybersecurity. Working in close proximity to Phil was a masterclass in turning challenges into opportunities. I learned how to build cross-functional teams with PhDs, engineers, and project managers to build a proof of concept (POC). I quickly understood how to take these POCs to client-facing teams and commercialize them to generate real revenue.

My team, the Office of the CIO within Engineering, had an incredibly cross-functional mandate. On one hand, we executed the commercial vision of engineering across technology commercialization (and open source), strategic investment, and industry consortia partnerships. We also worked directly with Investment Banking on priority TMT clients who, 90% of the time, were already strategic tech vendors of the firm. This was very much in line with the “One GS” operating ethos. 

My team members and I took a vertical focus based upon our demonstrated interests, with mine being cybersecurity and “distributed ledger technology,” something most people now recognize as a generalized term for blockchains.

I had a unique opportunity to work with Michael Aiello, the CISO at GS’s consumer bank at the time. We had a shared vision of a “single sign-on” for all consumer banks, similar to how you can use your Google account to login to most online platforms. GS leadership was intrigued as this would reduce technological barriers for consumers. We called it ‘FinID’ and developed a methodology to securely handle the credentials and personal data that protected both the user and bank platform from fraud. This was just a thought exercise at the time, but was a perfect example of the “intrapreneurial” mindset development available at GS.

Simply put, my role at GS helped me develop the muscle memory to scale a startup, while also understanding how to navigate the complexity of selling into highly regulated banks. I learned how to work with regulators in areas of fast-moving development. Given cybersecurity’s ever-evolving threat landscape, the technology will always move faster than the regulators move to protect us. 

Beyond this, I also gained insight on how to collaborate with competitors to lead industry best practices and set standards. I learned how banks can successfully and unsuccessfully collaborate with early-stage startups. In fact, Phil taught me that the best time to engage a vendor was early on, when GS could contribute to a startup’s roadmap and thereby advance its technology by years.

3. In 2021, about four years after you left the firm, you founded Reality Defender, which is a leading provider of deepfake detection. Tell us about what inspired you to start the business and about your journey until now.
My passion can be singularly defined as all things cybersecurity — from building consumer privacy solutions as a Google Intern, to launching the first privacy-focused e-voting blockchain at PlaceAVote, to my role at GS spearheading collaborative work to support cybersecurity commercialization and strategic investment.

During my time at GS, the bank was rolling out various consumer banking products that would fundamentally require new cybersecurity frameworks to protect consumer clients at scale. One of those included voice authentication, which was in its infancy. After battle-testing various vendors, we found that sometimes the vendor solutions would mistakenly recognize a computer-generated voice as a real voice. This was years before AI-generated deepfakes, so the primary test was on software used by the entertainment and gaming industry to create computer generated avatars and virtual humans. 

I learned that cybersecurity is not “if”, but “when”. At GS, Phil Venables pushed us to always think two steps ahead. “What if” spoofing technologies advanced and would become available to fraudsters and hackers? What could GS do now to prepare for this future? What technologies, processes, and policies can protect GS’s clients from these very asymmetric risks?

This was a turning point for me that sparked one of many insights that would set me on the path to develop Reality Defender.

I am constantly thinking about how Reality Defender can benefit from banks each time we receive an emergency phone call from a bank CISO looking for a deepfake solution. These banks are our partners as much as they are our customers, which means that honesty and transparency are paramount, especially when the threat landscape is changing in real time and we are ever expanding our “labs features” to allow banks to test-drive early products and features. It also means that we are accurate to the decimal when sharing our efficacy on accuracy of our deepfake detection. 

It takes years to earn trust, and mere minutes to lose it.

4. Reality Defender was recently named the winner of the annual RSAC Innovation Sandbox contest and “Most Innovative Startup 2024” for helping enterprises and governments detect deepfakes. What is your mission behind the business and how did you achieve this title?
Winning “Most Innovative Startup” at the RSAC Conference was a team effort and the culmination of many years of innovative research and commercialization. Together with our team of 40 (two-thirds of which are PhD researchers and engineers), we developed a framework to quickly research and launch solutions in design partnership with our clients. 

Reality Defender’s mission has always been to build the best deepfake detection solution and make it useful toward every-day cyber fraud professionals. We are thrilled to have received this recognition and we look forward to continuing to innovate and make a positive impact in the field of cybersecurity and AI research. Our clients and partners, who never waver with their support and trust in us, are also responsible for this monumental win. We look forward to expanding these relationships as we grow and work with key partners to stay ahead of the evolving threat landscape.


5. What was one of the hardest things you learned while building your own business?
The hardest thing that I have learned is saying no to a customer. As a founder, I am trained to always say yes because every conversation is a potential customer sale. But this is a fallacy. I believe that if you say yes to everything, you will excel at nothing. 

Over the years, Reality Defender has said no (respectfully) to dozens of customer revenue opportunities that are not within our mandate. Only by staying narrowly focused on deepfake detection, were we able to be first in market and consistently win research awards against much larger organizations.


6. As you look to the future, what are your ambitions for Reality Defender over the next few years?
Our goal is to develop the most robust deepfake detection solution to power proactive and accurate anti-fraud solutions against deepfakes, all while maintaining our ethics in sourcing data and training our models. 90% of all future media and communications will leverage generative AI and Reality Defender will be there to help countries, companies and consumers tell fact from fiction at crucial times.


7. Have you leveraged your GS network since leaving the firm? If so, how?
GS and its alumni network have proven invaluable since day zero. Various GS senior leaders have gone to bat for me with investors, partners, and regulatory paths. 

I was recently asked to give testimony to the US Senate and naturally turned to my GS contacts for advice. Within hours, the firm’s Office of Government Affairs in Washington D.C. reached out to help me navigate and prepare for what became a turning point in the push for Senate and Congress-mandated deepfake detection regulation. This was a full-circle moment for me and shows both the strength of my GS network and the firm’s dedication to its alumni.