Deepfakes: Battle for trust

Winner of the 2019 AI pitch battle and CEO of PHOTOCERT, Pasquale Saviano is looking to bring trust to a digital era deeply embroiled in deep fakes – He explains why prevention is better than cure

Congratulations on winning the pitch battle at the Malta AIBC Summit in November. There were some great start-up’s competing this year, what made you stand out from the crowd?

I like the fact that you call it a Pitch battle, but the battle is not only among start-ups but also, in a sense, with the crowd. Nowadays, everybody is online all the time and if you want to get your message across it has to be more interesting than the next WhatsApp notification. In my case, it probably was a mix of product novelty, current relevance (deep fakes) and fun facts.

Living in the digital era also means dealing with deep fake methods of image and video manipulation, how is PHOTOCERT endeavouring to protect authenticity?

Visual manipulation is like a bacteria that is continuously evolving. Today, there is no available cure in the market. We have developed several antibiotics, and we keep on researching new ones, but they cannot treat all the patients. For us, the best way of solving the problem is prevention. That is why we invented a process that creates a unique fingerprint that is attached to photos and videos as they are taken. This way we can be 100% certain of their authenticity.

Can you explain how you integrate blockchain technology into the PHOTOCERT tool?

Imagine an old man sitting on a chair writing a long string of characters on a ledger with a quill. The old man has to be sure each string is uniquely associated with a picture or a video. We are an old man, and as you probably know, the ledger is the blockchain.

Looking to the future, what’s in store for PHOTOCERT; in both the short and long term?

We have been able to find traction in the insurance (self-claim process, self-onboarding) and asset finance (self-inspections) sectors, and we are looking in the short/medium term to enter new markets; we are also working with computer vision and deep learning algorithms to spot “invisible” traces in pictures. In the long term our goal is clear: to make people trust what they see with their eyes again.

Beyond creating trust, PHOTOCERT can also impact a company’s ability to respond to interactions and automate the administrative process. How influential is this to the decision making procress for companies and individuals? 

This is quite important for our customers. For example, in the case of insurance companies, we not only prevent fraud but also enhance their claim processes, reducing costs and improving their customer experience. We have been able to do this by understanding their problems and looking for partnerships that leverage our product. For instance, we can use certified pictures to estimate car damages or to determine the type and value of an item making the onboarding process of content insurance simple.

How straightforward is it for users to integrate PHOTOCERT’S APIs into their app, web platforms and legacy systems?

It is quite simple, even for legacy systems. When we hire software developers, part of our recruitment process consists of giving them our APIs and asking them to integrate Photocert into an existing APP. It takes them from 2 to 4 hours to do so.

Our development philosophy is to create products that can be used in any system or application without requiring much effort from our customers. We value our time and we value our customers time even more.

What have been some of the challanges you’ve faced and how have you overcome them?

Our biggest challenge was to find our first big client. It was difficult because start-ups always need to show that they are growing fast if they want to appear attractive to investors. The answer to this problem belongs to one word that a green character from the Star Wars saga taught me: patience.

When you have a new product it is always difficult to find the first big customer. For us it was even more difficult because we have a B2B product designed for corporations, but we knew we had a great product – so it just came down to perseverance and patience.

Read more articles like this on Block magazine, view our digital version here.