Cyber Defense and Artificial Intelligence Paper Example

A debate on the impact of Artificial Intelligence (AI) on Human Values took place at the Peace Palace in the Hague as part of the 22nd European Conference on Artificial Intelligence(2016), and the first report of the 100 Year Study on Artificial Intelligence from Stanford University was published (Stone 2016). Both presented message of positive expectations while emphatically prescribing that moral, protection and security implications must be routed to guarantee, that the advantages of AI innovations will be spread comprehensively and reasonably. Ongoing advances in AI, are to a great extent because of the development and investigation of substantial data collections empowered by the internet, advances intangible innovations and all the more as of late, uses of "profound learning."

Is your time best spent reading someone else’s essay? Get a 100% original essay FROM A CERTIFIED WRITER!

Save 25%

In the coming years, as the general population experiences new AI applications in areas, for example, transportation and social insurance, they should be presented in manners that fabricate trust and comprehension and regard human and social equality. As opposed to being a risk to our reality or plotting to assume control over the rule of the world, in the coming decades we can anticipate that AI frameworks will progressively be connected in transportation, benefit robots, health care, education, public safety, and security. The ramifications of these advances in AI for occupations, social associations, security, and war were discussed among analysts and strategy producers in The Hague. With the new technology development in many areas, threats have come up concerning the Cyber Security of information stored in many organizations.

Cybersecurity is the protection of systems in the organizations, their data, and network in the cyberspace (Cyber Security Products and Services, 2016). Cybersecurity is a critical issue for many businesses. In a business or any organization, there are different threats associated with their systems, data, and networks. The threats include cyber-crime, cyber war, and cyber terror (Cyber Security Products and Services, 2016). Cybersecurity is a major problem, in many nations around the globe, research needs to be done concerning the possible measures to mitigate the problem (EBSCO, 2016).

With the increase in the number of cyber-attacks, more than human intercession is required for opportune assault investigation and suitable reaction. Evidence shows that the majority of the cyber-attacks are executed by intelligent agents like pc viruses. As such, the need to respond to such attacks using sophisticated agents with the ability to identify, evaluate, and offer the appropriate response has now become a key consideration. Stytz and Lichtblau (2005) express that these intelligent response agents should be adequately equipped to handle the entire operation of assault-reaction in the most convenient way. In other words, they should be able to identify the nature of the attack, its objectives, the best reactive measure, and how to prevent such attacks in the future. Besides, cyber intrusions are a worldwide hazard that presents a risk to any computer framework at an alarming rate. In the past, only the programmers could carry out digital crimes. However, that has changed in the contemporary world especially with the growth of the Internet where nearly anybody can access the learning and apparatuses for perpetrating these violations. Conventionally fixed algorithms have turned out to be ineffectual against fighting powerfully developing cyber-attacks. Helano and Nogueira (2006) 's diary discovered that cyber-attacks are the reason why inventive approaches are needed. Such approaches include the application of techniques for Artificial Intelligence (AI) that introduce strengths such as adaptation and learning to the profession of coding hence enabling developers to deal with cyber-crimes. Currently, many computing methods of AI (for example, Computational Intelligence, Neural Networks, Intelligent Agents, Artificial Immune Systems, Machine Learning, Data Mining, Pattern Recognition, Fuzzy Logic, and Heuristics) have been extensively implemented to help in cyber-crime identification and aversion. Indeed, AI enables the developers to arrange independent computing techniques with the capacity to adapt to their environment, apply autonomous programming methods, and autonomously carry out such processes as tune themselves, setup, analysis, and resolution. Undoubtedly, the AI systems are considered by many in the IT sector (Wang, Yang, Li, & Liu, 2008) as a promising niche with vast potential to address the issues related with data security and internet security.

Over time, the amounts of data stored and processed by computers and supercomputers have increased drastically, especially with the increase in the usage of internet to carry out business, work, communicate, socialize, and share information. As such, language barriers and geographical boundaries have vanished and the new virtual world seems to be overly populated than ever before. Dijle and Dogan (2011) expressed that the idea of crime is always present when people are involved, and therefore the internet has become intricately connected to the concept of crime and criminals. Brenner (2010) argues that a large portion of the cybercrime we see today essentially explains a new generation of certifiable crime whereby offenders use the internet as the primary tool to carry out old violations in new ways.

Prior to being known as machine intelligence, Tyugu (2011) illustrates that AI was started as an investigation discipline at the Dartmouth College in 1956. In the same way, Wang Yang, Li, and Liu (2008) depicted Artificial Intelligence in two ways. On the one hand, they showed AI as a science that seeks to understand the essence of knowledge and develop intelligent machines. On the other hand, they depicted AI as a discipline of study aimed at discovering techniques for resolving complex challenges that cannot be comprehended without making the correct decisions based on vast amounts of data. While using AI to provide cybersecurity, developers rely more on the second definition than the first one. In fact, AI research enthusiasts have been noted to incorporate approaches that make computers assume the intelligent human behavior manifested through thinking, learning, reasoning, and planning.

The whole topic of mimicking intelligence has been simplified to specific sub-topics that have unique traits which an intelligent system should manifest. Norvig (2003) illustrates traits that are mostly considered and include deduction, thinking, and reasoning through epitomized agents, neural systems and factual ways to deal with AI. Secondly, knowledge portrayal or ontologies. There is likewise multi-agent arranging and cooperation. Others are machine learning and data recovery through data mining and machine interpretation (Natural Language Processing). Also, Perception which involves object recognition, facial recognition, and object acknowledgment. Social Intelligence through sympathy reproduction would likewise get consideration. At last, Creativity, that is artificial intuition and artificial imagination.

A few strategies for fighting immerging digital security dangers were prescribed by The 2018 Cisco Annual Cybersecurity Report. Organizations were relied upon to expand their utilization of encryption. Nonetheless, to keep up, they would need to consolidate advanced tools - for example, AI - to counteract, distinguish and remediate potential dangers. Security experts were relied upon to spend more on tools that utilize AI and machine learning, which would help with the additional outstanding burden caused by the expanding danger of an assault, and enhance resistance. Since malware is regularly communicated inside scrambled web traffic, and sensitive information sent through Cloud frameworks, pertinent tools should have been set up to distinguish and keep the utilization of encryption for masking malicious action. Cisco (2018) 's report concluded that, after some time, AI ought to have the capacity to figure out how to naturally identify unusual patterns in encoded web traffic and Internet of things (IoT) conditions. Eventually, this would help enhance organize security defenses. Another enormous digital security issue has been the skill gaps: associations have not possessed the capacity to discover staff with the vital aptitudes. Artificial intelligence and machine learning devices would help beat these gaps.

This far, AI appears to have a place in the advanced security industry. Barclays Africa uses AI and machine learning to recognize and respond to cyber security threats. The CSO Kristen Davies expressed: "As the overall peril scene is advancing quickly, both in capacity and facilitated exertion on the assailant side, we ought to a great degree use advanced tools and developments to extend past the hazard themselves." The Cisco report addressed chief information security officers (CISOs), who said that they rushed to use AI and machine learning gadgets and that their security structure was developing in "headway and knowledge". Regardless, the officers say, the one issue with AI is the possibly high number of false positives, as this opposes the reason for reducing the outstanding burden. Certainty was, with time, this number ought to decrease as the machines get understanding.

Despite the way that AI is rapidly developing, affiliations are so far not shielded from the threat of a cyber-attacks. Before future protocols are proposed and executed in Cyber Defense by AI, we should have a predominant search at existing AI algorithm strategies. It is difficult to attempt to give a pretty much total review of all basically useful AI techniques in a concise overview. Instead, the procedures and plans have been assembled in the following categories as discussed below.

The history of neural nets dates back to 1957 where Frank Rosenblatt developed perceptron. Rosenblatt in his book "The Perceptron" demonstrates that just a few perceptrons joined together have the ability to learn and resolve complex challenges. Be that as it may, it is important to note that neural nets can many replica neurons. Thus, neural nets the advantage of colossally parallel learning and decision making. One of their most preferred traits is the speed of movement. Yang (2006)'s studies have shown that neural nets are effective in recognizing interruptions in the system and executing counteractive actions. Their efficiency has seen them recommended for application in DoS acknowledgment, computer worm identification, spam disclosure, zombie identification, malware gathering and in a forensic examination (G. Klein, 2010). Based on that information, it might be true to say that neural nets are used in cyber security because of their speed, either in hardware or used in graphical processors.

Conclusion

Expert systems are evidently the most used part of AI apparatuses. Kivimaa, Ojamaa, and Tyugu (2009) 's notes portray the Expert System as a component designed to find answers to inquiries in some application's space introduced either by a client or by another product. Typically, a specialist framework entails a knowledge database, which stores information about a given application. Besides an information base, an expert system fuses a derivation motor for deciding answers reliant on this data and, possibly, additional learning about a situation. It might be particularly used to make decisions like in therapeutic analysis, in finance or on the web. Currently, there is an unimaginable collection of expert systems from minimal particular decisive systems to considerable and present-day hybrid structures for dealing with complex issues. The Case of Cyber Defense Expert systems is one of security planning and it empowers astonis...