Connecting Linkedin...


Using AI to Benefit Cybersecurity

13 days ago by Andrea Amato

Cybersecurity has been a growing concern exacerbated by the Covid-19 pandemic. Cybercriminals have implemented novel strategies to phish users into donating toward pandemic efforts and other similar philanthropic actions. It can be difficult to intervene on security threats through human intervention alone, which is where we’re far more equipped to do so using AI technologies. Although imperfect, we’ve a fighting chance in improving our cybersecurity efforts that surpasses our own human abilities.

Specifically, AI and machine learning (ML) have re-structured security and IT jobs in quickly reading high volumes of data to determine any found information threats. Such technologies are also able to distinguish between data, including malware, that could either represent a phishing attack or a malicious download. Like all ML approaches, trained algorithms only become more intelligent in deciphering between what constitutes a security threat.


Advantages of AI

Whilst AI represents an umbrella term that encompasses numerous kinds of technologies, these often interlink to produce solutions for complex security issues. The main benefits in using AI for cybersecurity purposes are threefold:

  • Regular learning: over time, AI becomes more sophisticated in identifying cybersecurity threats and risks in scanning copious quantities of data.

  • Differences between threats: AI can explore the relationships between different kinds of threats and suspicious behaviour in a matter of seconds.

  • Time-effective: AI can produce a thorough risk and security analysis in quick time, leading security teams informed decision recommendations with immediacy.

In a broader sense, AI can support security and vulnerability management in protecting an organisation’s network. Since Covid-19, cybercrime has increased to 300%, whereby 53% of adults believe remote work has made their security more vulnerable (Norton, 2021). As AI can assess data systems at a much quicker pace than staff, it’s worth implementing some of these technologies in focusing on security tasks.

Whilst security threats can differ daily and cybercriminals innovate creative means to scam users, AI is less affected by these. This is because in exploring these threats, AI is able to compute through the basic elements of a security threat, no matter the accompanying narrative. This means that AI can regularly assess security threats with little human intervention.

As cybersecurity grows more sophisticated, especially with work from home jobs that is more isolated from dedicated data teams, AI will become an integral part of an organisation’s security agenda.


How Does AI Work?

Albeit the intricacies of AI are beyond the scope of this article, it’s worth understanding its main components in order to apply the technologies in a cybersecurity context. Presently, AI informs our decision-making largely through assisted intelligence, which is based on improving our current methods as individuals and organisations. Augmented intelligence, assisted intelligence’s superior, is an emerging technology for its core works beyond human capabilities. Lastly, albeit not yet mastered, autonomous intelligence refers to technology that can act with no human intervention (self-driving cars, for example).

AI works on the premise that it adopts and utilises human intelligence to infer its actions. Human qualities such as learning from mistakes and the ability to acquire new knowledge forms the basis of ML, deep learning, and neural networks:

  • Machine learning, through statistical analysis, learns to use data over time and not via human programming. The more targeted ML is, the better the performance.

  • Deep learning, a subfield of ML, works by learning data representations and not task-specific algorithms. It requires less human intervention than ML does and is today used to work on autonomous intelligence projects.

  • Neural networks allow computers to learn from observational data. Although this kind of AI works through its own sophisticated measures (including weights, nodes, etcetera), it’s also related to machine and deep learning strategies.

The above are some examples of how we use AI to perform tasks that would otherwise be highly time-consuming for humans. As we continue to develop these technologies, their importance and our reliance will make for its integration in our everyday jobs in Malta and elsewhere more inevitable.


Applications in Cybersecurity

As AI continues to evolve with little human intervention necessary (albeit this is not the case in terms of maintenance and testing procedures), its technologies will grow better suited to meet cybersecurity demands and risk mitigation. Security is a complex phenomenon for all organisations to tackle, especially for small-scale organisations with little IT resources.

Whilst the impressive little time spent on threat detection provided by AI is certainly helpful, cybersecurity has presented particular challenges that AI will have to outsmart. Cybersecurity threats reach a great number of devices, sometimes targeted, and does so through several mediums (emails are dominant, but mobile phones are increasingly becoming targeted). Albeit widespread recommendations for organisations to include dedicated IT teams to manage such security threats, many companies simply do not possess the resources to do so. In this way, the number of threats overwhelmingly surpass the available staff.

It’s for this reason AI has stepped in to overcome the sheer quantity of cybersecurity concerns. Organisations already implementing and integrating these technologies into their IT jobs and systems are helping to develop smart algorithms that can be applied on a global scale. There are currently several means AI are assisting humans in the fight against security breaches:

  • Access into devices: AI can monitor workplace devices over a common network, identifying threats as these arise and flagging them to the appropriate teams available.

  • Informed decision-making: AI can build a sophisticated understanding of workplace trends that will be used by cybercriminals to create new threats, and tackle these quickly, regardless of the latest trending security crimes.

  • Improve current systems: working alongside other technologies, AI can improve present IT jobs and systems in place and identify whether these require updating or further performance improvements, strengthening an organisations security system overall.

  • Regular risk checking: even when security threats are dormant, AI can work to analyse and assess holes in security systems that leave room for easy cybercrime access. It can also pose recommendations to organisations in how these issues can be solved.

Overall, AI can be used for multiple purposes to support IT jobs in Malta, beyond, and in data teams. Be it specific, targeted tasks or general security protection, these technologies will help strengthen the security of an organisation generally. Many large-scale companies currently make use of AI for these purposes, such as Google and IBM.

Nevertheless, AI is not perfect, and is regularly criticised for its numerous biases and ethical concerns, which creates several challenges for jobs in tech. For any organisations interested in implementing AI, it’s worth exploring these challenges and solutions before doing so. In this way, the field of AI can grow to become more accessible and reliable for organisations of any scale.