Artificial intelligence, a double-edged sword.

The Artificial Intelligence he was a big protagonist at the fair international technology CES, held this week in Las Vegas: from toothbrushes, grills to cars now integrate this technology in one way or another. Considering the spread of its use, Kaspersky experts conducted an analysis on the major impact of artificial intelligence on cybersecurity, exploring its implications and focusing on the uses that defenders and regulators, as well as cybercriminals, can put to this tool.

Kaspersky investigation details how threats related to artificial intelligence will evolve and reiterates the importance of having a regulatory framework that enables the responsible use of this tool. The most relevant aspects of the study "The impact of artificial intelligence on cyber security" are below:

More complex vulnerabilities As Large Scale Language Models (LLM) they are integrated into products that are more consumer oriented, new
complex vulnerabilities at the intersection of probabilistic generative artificial intelligence and traditional deterministic technologies, expanding the attack surface that cybersecurity professionals must protect against. This will require developers to explore new security measures, such as user approval for actions triggered by LLM agents, which are essentially AI systems for understanding and generating text in a similar way to humans.

A comprehensive AI assistant for IT professionals cyber security. Researchers and members of the Red Team They use generative artificial intelligence to create innovative cybersecurity tools, potentially leading to an LLM or machine learning (ML) assistant.

This tool could automate tasks The Red Team, offers order-based leadership it is executed in a pentesting environment (a prevention exercise in which a cyber security expert tries to find and exploit vulnerabilities in a computer system).

Neural networks will increasingly be used to generate images to aid fraud. This year, fraudsters could expand their tactics using neural networks to create more convincing fake content using artificial intelligence tools. With the ability to effortlessly generate convincing images and videos, malicious actors pose a greater risk of escalation cyber threats related to scams and fraud.

AI will not become the driver of revolutionary changes in outlook for threats during 2024. Despite Based on previous trends, Kaspersky experts remain skeptical that artificial intelligence will significantly change the threat landscape anytime soon. While cybercriminals will adopt generative artificial intelligence, so will cyber defenders, who will use the same tools, or even more advanced ones, to test security improvements in software and networks, making them unlikely to drastically change the outlook for attacks.

Improve related regulatory initiatives with AI with the contribution of the private sector. With its rapid development, technology has become a matter of policy formulation and regulation. The number of regulatory initiatives related to artificial intelligence is expected to increase. Non-state actors, such as technology companies, given their experience in the development and use of artificial intelligence, can provide invaluable knowledge for discussions on regulation of AI on platforms, both global and national.
Watermarking for AI-Generated Content More regulations will be required, as will service provider policies to mark or identify synthetic content, and service providers must continue to invest in detection technologies.

Developers and researchers, for their part, will contribute to the creation of watermarking methods for facilitate identification and provenance synthetic media.

“Artificial intelligence in cyber security is a double-edged sword. Its customization capabilities strengthen our defenses and provide a proactive shield against evolving threats. However, this same dynamic presents risks, because Attackers use artificial intelligence to design more complex attacks. Striking the right balance, ensuring responsible use without sharing too much sensitive data, is key to protecting our digital borders,” he comments. Vladislav Tushkanov, security expert at Kaspersky.

This analysis extensive, which is part of the Kaspersky Security Bulletin (KSB), is annual compilation of detailed forecasts and reports that illustrate key changes in the dynamic field of cybersecurity. You can view the complete study here.