Chapter Ten of seventeen
Chapter 10: A.I. & Cyber-security
Worldwide, according to Statista, spending on Cyber-security is forecasted to reach $650 billion by 2030. One of the most under-researched, under-invested, and underappreciated aspects of Artificial Intelligence, is protecting it. But protecting it from whom? You have many outside threats. But what if the threat comes from within? These are questions that few in human society have pondered, but these questions that must be addressed as the Coronavirus, and economic woes, accelerated automation and A.I. adoption the globe over.
Cyber-security is a moving and dynamic landscape. According to IBM's Chairman and CEO, "Cybercrime is the greatest threat to every company in the world." In order to understand the importance of Cyber-security, let us first explore a number of statistics, provided by our friends at webarxsecurity.com, that do a great job explaining how common and problematic hacking is on daily digital life:
1. Cyberattacks:
"A cyberattack occurs every 11 seconds, and a ransomware attack every 2 seconds." (IBM X-Force Threat Intelligence Index 2023)
2. Data breaches:
"In 2023, over 6 billion records were exposed in publicly disclosed data breaches worldwide." (IT Governance UK Blog, November 2023)
"The average per-record cost of a data breach is $165, a 1 dollar increase from 2022." (IBM Cost of a Data Breach Report 2023)
3. Cybersecurity effectiveness:
"67% of security professionals believe existing security controls are insufficient to fully protect against cyberattacks." (EY Global Information Security Survey 2023)
4. Malware:
"McAfee Labs identified over 120 million new malware samples in Q3 2023 alone." (McAfee Labs Threats Report Q3 2023)
5. Website attacks:
"A study by Positive Technologies found that attackers breached over 44,000 websites in the first half of 2023." (Positive Technologies Global Security Report H1 2023)
6. Phishing and social engineering:
"The Verizon 2023 Data Breach Investigations Report found that phishing remained the most common attack vector in 82% of breaches." (Verizon Data Breach Investigations Report 2023)
7. Breach motivations:
"The IBM Cost of a Data Breach Report 2023 states that 82% of data breaches were financially motivated, with espionage at 9%." (IBM Cost of a Data Breach Report 2023)
8. Malicious email attachments:
"Microsoft's 2023 Digital Defense Report revealed that malicious Office documents continue to be a major threat, accounting for 44% of email-borne attacks." (Microsoft Digital Defense Report 2023)
9. Password usage:
"A Google Cybersecurity Whitepaper published in 2023 estimates that there are now over 84 billion active passwords used globally." (Google Cybersecurity Whitepaper: The State of Passwords 2023)
10. Major data breaches:
"As of October 2023, the T-Mobile data breach affecting 54 million customers is the largest reported breach of 2023." (TechCrunch, October 2023)
11. Public awareness of data breaches:
"A recent Ponemon Institute study found that 58% of Americans have never checked to see if they were affected by a data breach." (Ponemon Institute: 2023 Privacy Pulse Report)
12. Data breach cost:
"The IBM Cost of a Data Breach Report 2023 found that the global average cost of a data breach increased to $4.45 million, a 15% increase over 3 years." (IBM Cost of a Data Breach Report 2023)
As you can see, the impact of cyber-security is beyond profound on the global digital infrastructure. Additionally, the problematic nature of cyber-security and the ineffectiveness of current applications and methodologies to adequately protect current infrastructures, has led to the integration of A.I. into Cyber-security systems to provide the compute intensive analytics that are impossible for a human to detect, due to the sheer number of breaches, vulnerabilities, and evolving nature of the threat. However, the ability to use machine-learning and deep learning by hackers for nefarious purposes has also become a reality. Therefore, although the defenses have evolved, so has the threat. Now, let us discuss the different penetration barriers, in the defense of a mock network.
The first line of defense are people. Whether you are a sole-proprietor, or a Fortune 500 company, you can have the best defensive Cyber-security system in the world, but if you people do the wrong things, it will become money wasted. Suppose that I wanted to get into your network. What might I do? I could try to hack in from the outside, but that's so much work! What if I walked into you office, and pretended to want to make an appointment, but needed to access my email first. I ask your secretary for the network name and WiFi password, and Voila! I'm in. Maybe it not a person asking your secretary for access in-person, but maybe I am sending an email with a malicious attachment or link? As a result, your people are as important of a defense as a million dollar cyber-security system. A.I. will be no different when trying to select the best target to penetrate an enterprise or consumer system.
The second line of defense is your password and username authentication system. Is your password secure? Do you change your password frequently? Do you use two-factor authentication? Do you have a separate login and password for a guest network? If an A.I. or hacker has access to enough computing power, they can crack your password. That's why a two-factor authentication system is critical to protecting your infrastructure. Other methods such as bio-metric scans, eye scans, and finger scans are beginning to replace legacy systems. However, if a person is kidnapped or killed, their identify is still at risk. New solutions and methods will be needed for this second line of defense, but for now let us move on to the third.
The third line of defense is the system firewall. Many consumers have their firewall turned off, either because it is annoying, or because they have expired anti-virus software on their system. An active fire-wall is standard within enterprise systems. However, security also depends on the system. Most consumers have a Windows operating system, and therefore, most hackers are well-versed on the vulnerabilities of this system. Switching to a Linux-based system is much more secure. However, A.I. is utilized primarily on Linux-based systems, and thus any A.I. attacking would likely be the most proficient on its own system. A.I. can assist currently in the dynamic patching of holes in the firewall, where human intervention would be too slow and inaccurate. Likewise, malicious A.I. would have an amazing ability to infiltrate hidden vulnerabilities and gaps in firewall cyber-security systems. Let us observe the the fourth layer.
The fourth layer of cyber-security is the anti-virus software. For Windows-based systems, Norton or McAfee are popular anti-virus software applications. On Linux-based systems, chkrootkit or rkhunter are popular cyber-security applications for scanning and cleaning systems of viral, malware, and rootkit threats. What's interesting about anti-virus software, is that some must be installed as soon as it is purchased, other software can be installed after the initial use. The reason is simple. Some computer viruses can hide within your system and prevent the anti-virus software from cleaning the drive. It is a best practice to install anti-virus software when a system is first used. It remains to be seen how A.I. will manipulate anti-virus software, but there are current examples. Let us look at three examples, of .A.I.-powered anti-virus software.
The first examples comes from IBM's Watson. IBM QRadar Advisor with Watson, leverages the power of cognitive A.I. to automatically investigate indicators of compromise and gain critical insights. QRadar consolidates log events and network flow data from thousands of devices, endpoints and applications, correlating them into single alerts -- so you can accelerate incident analysis and remediation.
The second example comes from Google's Gmail, which uses machine learning to block 100 million spams in a day.
The third example comes from Nvidia's DeepStream Software Development Kit, which allows for the building and deploying of A.I.-powered Intelligent Video Analytics apps and services on edge-computing devices, such as the Jetson Xavier, or in the cloud.
The fifth and last layer of cyber-security involve the securing of physical backups of user data. Currently, the most effective way to secure information, is to save it do a drive and then take that drive offline, disconnecting it from the computer, internet, and reach of hackers. This does not mean it is totally, safe, as it could then be lost or someone who really want it, can break in and take it.
An attempt at a solution to this problem, involves the use of cloud-based storage, that automatically backs up user data in the cloud. The issue with this, is that the cloud-based storage is susceptible to hacks, and even if your data is not being targeted, your data and web-based services could still be impacted due to the intermingling of networking involved in administering the cloud. For example, in October of 2019, Amazon's S3 or Simple Storage Service, was attacked taking down thousands of websites. Now, let us explore a new proposal to ensure future security for A.I.-based Cyber-security attacks.
I am proposing a sixth line of defense, specifically targeted at A.I.-powered cyber-security threats. This defense consists of Black Bots, that are controlled by logic, and when the system does something that is illogical, such as self-harm, the unauthorized opening of ports, or manipulation of data, will cause sleeping cells of Black Bots to awaken and spring into action. These bots are analogous to white blood cells protecting a host, except these are Bots protecting a network and its data contained therein. These sleeping cells of bots will trigger trap doors, isolate viral system infections, while also preventing the virus or malicious A.I. threat from communicating with other A.I. outside of the system. At this point, the bots will have the ability and capability to delete the corrupt data from the hard drive, by attaching itself to the infected host similar to a glycoprotein, injecting the host with its own digital signature until the virus is benign and no longer a virus. Then the virus can then be deleted, without fear of further spread or viral reignition.
Cyber-security will come to play an even more important role as we automate everything in sight, and autonomous vehicles become standard in society, rather than something far-sighted. All of these machines will connect with everything else in a Smart City or Smart Home. These systems must be multi-layered, and provide for adaptability to the evolving threat of new and increasingly complex attacks. The increase in connected devices, as a result of the internet of things, will only increase the openings for attacks, and increase the complexity of the layering of each secured system. But also, remember, that irregardless of how sophisticated your system is, it is only as sophisticated as your people. In the next chapter, we will discuss A.I. and the programming languages that are necessary to develop it.
Exercises
Can you or your team explain what Cyber-security is? How does it apply to Artificial Intelligence?
What are some of the new tools that are available for Artificial Intelligence and Cyber-security?
How many layers of Cyber-security are there? Can you name them?
How will your project utilize Cyber-security to protect your data? Here is more concerning Artificial Intelligence and its impact on Cyber-security.
Ean Mikale, J.D., is an eight-time author with 11 years of experience in the AI industry. He serves as the Principal Engineer of Infinite 8 Industries, Inc., and is the IEEE Chair of the Hybrid Quantum-inspired Internet Protocol Industry Connections Group. He has initiated and directed his companies 7-year Nvidia Inception and Metropolis Partnerships. Mikale has created dozens of AI Assistants, many of which are currently in production. His clientele includes Fortune 500 Companies, Big Three Consulting Firms, and leading World Governments. He is a former graduate of IBM's Global Entrepreneur Program, AWS for Startups, Oracle for Startups, and Accelerate with Google. Finally, he is the creator of the World's First Hybrid Quantum Internet Layer, InfiNET. As an Industry Expert, he has also led coursework at Institutions, such as Columbia and MIT. Follow him on Linkedin, Instagram, and Facebook: @eanmikale