Mobeen Tahir, Director, Macroeconomic Research & Tactical Solutions, WisdomTree: How generative AI is changing the world, including cybersecurity.
In 2024, the inhabitants of 55 countries, representing the 42% of the world‘s population, will vote to elect their leaders. Among the various elections, there are those for the presidency of the United States and the Parliament of the European Union. In addition to attracting billions of people to the polls, these events could attract the attention of bad actors.
The electoral process at risk
With the generative artificial intelligence (AI) tools at their disposal, they seek to sabotage the electoral process directly by hacking the systems. Or to influence public opinion using disinformation or deepfakes, with the possibility of . How is generative AI transforming cybersecurity, and what might it mean for investors trying to predict the future of this investment theme?
How generative AI can help attackers
Historically, basic cybersecurity training was sufficient for people to identify and defend themselves against attacks phishing. Incorrect grammar, suspicious domains and questionable links were generally easy to spot. However, generative AI tools, such as WormGPT, trained on malware data and developed specifically for criminal activity, can eliminate easy-to-spot clues in phishing emails. Also lower the bar for those who wish to commit malicious acts.
What is polymorphic malware code
It can also encourage polymorphic malware code. That is, a type of program that learns and evolves to automatically become more intelligent after a failed attempt. In other words, if the target doesn’t update their security software, the malware code will come back stronger to exploit any vulnerabilities in the system.
The quality of deepfakes goes up
Generative AI is also improving the quality of deepfakes. According to the World Economic Forum, over 100 deepfake videos featuring ads by British Prime Minister Rishi Sunak were identified on Meta between December 9, 2023 and January 8, 2024. Many of these elicited emotional reactions using language such as “people are outraged.”
How generative AI can help defenders
Using generative artificial intelligence to defend against increasingly sophisticated attacks is an essential necessity. For example, this can help develop more advanced training modules to enable users to better defend themselves against high-quality phishing attacks. It can also be used to analyze large amounts of data to identify patterns, trends and anomalies.
These can reveal vulnerability into the system and help cybersecurity teams close gaps before polymorphic malware code makes its return. Additionally, it can be used to automate repetitive processes, streamlining tasks that are not only tedious, but also prone to human error. Such as incident response, threat detection and malware analysis.
Cyber security is more important than ever
The figures below are an alarming reminder that the importance of cybersecurity cannot be overstated. Most people who own a smartphone or laptop use dozens of cloud-based software applications. Criminals recognize this larger attack surface and are increasingly targeting their victims via the cloud.
Hitting through the cloud
There is also an alarming increase in sophisticated attacks without malware: 75% in 2023 compared to 40% in 2019. This suggests that attackers are switching to faster and more effective means of infiltrating target organizations. Using tools like social engineering instead of always relying on inserting malware into the victim’s system. This is an important reminder that cybersecurity training shouldn’t just teach users how to defend themselves against phishing attacks, but also how to stay on guard against physical deception.
Generative AI is changing the world, including cybersecurity
The evasion time of a cyber attack represents the time needed for a “hacker” to move laterally within an organization. It means gaining access to other users after compromising the first victim. The average time to evade a cyber attack dropped from 84 minutes in 2022. To 62 minutes in 2023, with the shortest time ever recorded being just over 2 minutes. Finally, the average cost of data breaches to organizations in 2023 was $4.45 million.
This means that the attackers they are increasingly faster, use a wide range of tools and cause serious damage to their victims. Fortunately, organizations are waking up to this reality. According to IBM, 84% of executives plan to prioritize generative AI-based cybersecurity solutions over traditional ones in 2024.
What this means for investors
Cybersecurity was one of the best performing themes in 2023. The WisdomTree Team Cybersecurity UCITS Index returned 66.5% in 2023. It even beat the NASDAQ CTA Artificial Intelligence Index, which returned 55.9%. It is obviously impossible to predict whether we will see similar numbers again in 2024. But if the market performance of thematic strategies is the result of strong positive currents in the underlying technologies, combined with a broader appreciation of those trends, there is certainly a lot to be excited about when it comes to generative AI. As well as its impact on cybersecurity.