Israeli startup Gnostic raises $11 million to prevent artificial intelligence from revealing sensitive information about organizations. Here's how it works
This week, the good news was finally announced: the company Knostic The Israeli company, which I have the honor of being a part of, managed to raise $11 million. And rightly so, because the service we provide can make any company safer.
It is customary to say that a mediocre future from the beginning of the twentieth century would have predicted that there would be cars on the roads. A brilliant future would have predicted traffic accidents and traffic jams – in other words, the consequences of technology when it reaches widespread implementation.
Two years ago, Gadi Evron and Sunil Yu had such a futuristic flash. Gadi and Sunil realized that all companies were about to implement artificial intelligence in their systems. This intelligence would help employees in a variety of ways. If an employee wanted to know, for example, whether a certain supplier was trying to cheat him with an especially high price quote, he could ask the artificial intelligence to find him similar price quotes from previous years, and it would do so and recommend what to do. If a human resources manager wanted to know what a particular employee's salary was, she wouldn't have to go through the trouble of searching for the right Excel file, but could simply ask - and the artificial intelligence would find exactly the right data for her in the appropriate file.
The problem begins when we realize that every company has information that it doesn't want to share with all employees. A human resources manager might know what other employees are paid, but we wouldn't want a programmer or an air conditioning technician to have access to that information. And of course, every company has millions of old files on its hard drives, which may contain a wealth of information that managers wouldn't want employees to have access to. But artificial intelligence wants to help so much that it might happily and graciously provide this information to employees.
These may sound like small problems, but these are exactly the kind of obstacles that can cause companies to avoid implementing artificial intelligence into their systems.
The two entrepreneurs understood all this and decided to found Nostic to build a solution to the problem. The solution itself is based – of course – on artificial intelligence, and it comes in two stages.
In the first phase, we explore all the ways in which you can 'work' the AI. Lure it, trick it, trick it, or any other word you like, as long as the end result is that the AI gives you secret corporate information. We develop a series of methods – you could almost call them "psychological" – to trick the corporate AI, and then we run them on the clients' computers, and watch the AIs there steal the organization's most secret files.
Why do we do this? Simple: because we assume that some malicious attacker, or just a curious employee, will do something similar in the future. We get ahead of the curve, and discover the organization's secret and problematic files ourselves.
Then we let another AI – the one we’re developing – read them. Why? To understand what they contain, and label them so that only the right employees can open them.
At the end of this process, which is applied to all of the company's databases, each file is given its own label. Files with confidential personal information become available only to HR managers, while files containing the company's secret strategy, or transcripts of senior management meetings, are designated as "top secret" and are only visible to members of management. All of this is done automatically, without a human eye ever looking at these files.
If humans were to do this work in a large company, it would take them years of reading files. And to protect the secrets in those files, the human taggers would eventually have to be buried inside the walls, as was the custom of the pharaohs and emperors of China. But with artificial intelligence? The process can be completed in a few days, and it doesn't tell anyone about what it read. The process should help even companies that don't yet want to implement artificial intelligence in their systems, but 'just' want to protect millions of files that have accumulated throughout their years of existence.
This is Nostic's value proposition, and last week it received a wonderful vote of confidence from investors. A group of more than a hundred senior VPs in the field of information security and cybersecurity chose to invest their best money in Nostic, along with several other serious investors.
Want to try the services of Gnostic? Talk to us.
Together, we will help all companies and organizations leap forward into the next great revolution, in the safest way possible.
More of the topic in Hayadan:
