SharkGate’s Artificial Intelligence (“Deep Sea”)

Solution: Part 3 of 3

Ecosystem – Part 3

The SharkGate Ecosystem protects websites against current and next-generation cyber threats. The core components of the Ecosystem are:

  • SharkGate Security Plugin For Websites (“SharkGate Plugin”)
  • SharkGate Website Threat Defence Database (“WTDD”)
  • SharkGate blockchain-based AI (“Deep Sea”)

These 3 components form the basis of our unique approach exclusively dedicated to protecting websites and constantly evolving the ecosystem for website cyber threat protection. This Ecosystem will finally provide the solution to protect websites against current and next-generation threats.

In this post we will describe the third of these parts “SharkGate’s Artificial Intelligence (‘Deep Sea’)”.

SharkGate’s AI (“Deep Sea”)


Big Data is not Smart Data, so for it to be useful for the ecosystem the SharkGate artificial intelligence named ‘Deep Sea’ is required.

The HTTP attack requests, hacker fingerprints and malicious files from one website are of incredible value to other websites that want to be protected from such future attacks. The SharkGate Plugins on each site provide this data to the SharkGate ecosystem. This consolidated data combined with existing data from the WTDD is processed by Deep Sea to learn from and improve the ‘Distributed Acquired Attack Immunity’ of the SharkGate ecosystem and subsequently the protection of every website on SharkNet.

Deep Sea uses distributed consensus when learning from the Big Data and improving protection assets (e.g. threat data classifications, firewall/scanner rules, hacker identifications and markings, score adjustments, known exploits, malicious files, adaptive rules, false positive identifications, malware signatures, etc.).

Deep Sea coordinates the memory of each attack encountered on any site worldwide and thus enables any site on SharkNet to mount a strong response if the attack is attempted again. It 17 also creates ‘adaptive’ rules that evolve during the lifetime of the network as an adaptation to a threat and prepares protection for future similar, but yet unseen, attacks.

Deep Sea ensures one of the key goals of SharkGate is met: A hacker attacking one website in the world that is part of the decentralized cybersecurity system actually strengthens the hacker protection of every other website on the network. Note: The term artificial intelligence (“AI”) used by SharkGate within the Ecosystem encompasses many fields including machine learning, pattern recognition, deep learning, neural networks, anomaly detection and more.

Deep Sea – Innate & Adaptive (Acquired) Immunity

The self learning protection process of Deep Sea has some analogies to the marvellous human immune system. It uses a system of innate and adaptive (acquired) rules. Acquired rules are produced by 2 types of “Distributed Attack Immunity” called “Artificial” and “Natural”.

Rules Types

  • Innate Immunity Rules – Rules that require no additional “training” to do their jobs. The plugin continually updates with these rules as they become available. Our first set of rules consisted of industry standard rules for each type of Website PLUS thousands of rules developed in-house over the last 4 years by the SharkGate firewall and OneHourSiteFix malware scanner. SharkGate and OneHourSiteFix are already keeping vast amounts of websites clean from infection, so even from day 1, these innate rules provide a very strong protection indeed.
  • Adaptive (Acquired) Immunity Rules – Produced by SharkGate Plugin’s AI and by the Blockchain AI ‘Deep Sea’ from contributed data on the network. These rules learn and improve upon exposure to attacks. The advantages of the adaptive rules are that they are able to adapt and protect from new types of attack as they emerge. These rules do not come into full force until they have gained the experience necessary for optimal attack and malware protection. The feedback of these rules to Deep Sea also enables further Innate rules to be created at a later date. Although the formation of global threat memory and experience from these rules occurs 24/7 throughout the life of the blockchain, it is expected the most rapid gain will be in the next few years of the ‘Deep Sea’ blockchain solution.

Immunity Creation Types

  • Natural (Acquired) Distributed Attack Immunity – This is a long-term active memory that is acquired by ‘Deep Sea’ based on attacks against any of the websites on the distributed network. With the ecosystem building the collective intelligence from each attack. it means all other sites worldwide are then equipped to mount a strong response if such attacks are detected again. This type of immunity is ‘adaptive’ because it occurs during the lifetime of the blockchain as an adaptation to attacks and prepares the cybersecurity system for such future challenges.
  • Artificially Acquired Distributed Attack Immunity – This is when threat payloads from selected nodes, oracle sources (e.g. external threat database, consensus agreed on attack scripts, etc.) are run against the system for it to learn from and improve the overall immunity. This works in the same way the active immunity of a human body would be generated artificially through vaccination.

Download Whitepaper

There is no denying there is a desperate need for a new way to protect sites from hackers. Think about how Google transformed the search economy. Think about how Uber and Airbnb transformed their sectors. SharkGate is now doing the same for website cybersecurity. Further information about the existing crisis and our solution to the problem can be found in our whitepaper.


Join Our Mailing List