The public’s ability to trust what they see published from media and governments is paramount to our modern global society. We believe that the AI tools available today for anyone to download and use are an inherent threat to the ongoing trust of any audio or video.
Tell us about yourself?
I have been a computer and AV club nerd for twenty years. My first job was working as an intern for a cybersecurity firm in the pacific northwest of the United States, performing penetration testing and vulnerability scans for small local banks. That is what sparked my interest in cryptography, encryption, and eventually lead me to research and understand blockchain.
My interest in combating deepfakes started shortly after I first learned about free software being available and viable for anyone to produce their own deepfake content several years ago. When I further learned how the major players working to fight deepfakes were approaching the problem, I knew that a novel approach would be needed.
Microsoft and Facebook were simply improving the quality of deepfakes produced by Generative Adversarial Networks (GANs).
If you could go back in time a year or two, what piece of advice would you give yourself?
Don’t waste six months before understanding the utility that NFTs can provide as a digital certificate of authenticity.
What problem does your business solve?
We believe that the AI tools available today for anyone to download and use are an inherent threat to the ongoing trust of any audio or video. Those same tools are also very clearly here to stay, as they have real utility when employed by non-malicious users.
The public’s ability to trust what they see published from media and governments is paramount to our modern global society.
Our solution provides a way for any company, government, or person to easily claim all their digital assets as NFTs on a decentralized system. By storing the content on a blockchain using advanced techniques, we can detect fraudulent media.
What is the inspiration behind your business?
I was inspired when I learned about the widely available, free software to produce deepfakes, and when I learned that the major players fighting deepfakes were all trying to create detection systems that would only serve to improve the quality of future deepfakes.
What is your magic sauce?
We are not creating an AI to detect deepfakes based on visual or audio artifacts/errors/glitches, because we understand that the net effect of an improved detection method is an improved GAN that can both 1) fool the detection AI and 2) be more convincing to human consumers of the content.
Rather, we are leveraging NFTs as a digital certificate of authenticity to ensure that content producers can claim the content they publish. In turn, this will provide a mechanism to help content producers dispute the veracity of deepfake content produced based on their legitimate published content as well as content that has been falsely generated out of whole cloth (and claims to come from the content maker).
What is the plan for the next 5 years? What do you want to achieve?
We are working with the US National Science Foundation on an SBIR grant for non-dilutive funding to cover the Technical and R&D costs with making this solution a reality. We hope to get the grant approved, and then leverage those funds to build a working solution that will protect digital multimedia evidence, such as police body camera footage, and ensure that there are no questions about its trustworthiness.
What is the biggest challenge you’ve faced so far?
I am dealing with technologies that are widely considered to be buzz words in the tech industry right now: blockchain, NFTs, deepfakes, and AI. Many times, depending on my audience, I have to be diplomatic about mentioning these key terms.
How do people get involved/buy into your vision?
We need to hear from content producers like podcasters, YouTubers, musicians, and more to understand what kinds of needs they have for a solution like this, and what the current best tool is for managing cross-posts of their content across social media channels.
Please go to https://DeepFakeStop.com/signup to let us know that you are interested.