How AI is changing the landscape of cybersecurity and cybercrime

The growing popularity of AI tools may pose a threat to cybersecurity

The fast improvement of AI capabilities is raising concerns

AI is one of the most revolutionary technologies of our time, and it is changing the way we think about cybersecurity and cybercrime. In just two months, the OpenAI-developed chatbot surpassed 100 million active users and established itself as the fastest-growing consumer application in history. Users looking to use AI to compose malware and emails for espionage, ransomware attacks, malicious spam, and other nefarious activities currently dominate discussions in cybercrime forums.

One cybercriminal shared the code for an information stealer they created using ChatGPT as a demonstration of their abilities. The malware, written in Python, was capable of locating, copying, and extracting 12 standard file formats from a compromised system, including Office documents, PDFs, and images.

Newly discovered campaigns even show how criminals teach artificial intelligence robots and assistants to inject malicious code into their programmed code to spread threats. Trojan Puzzle campaign shows that cybercriminals use the popularity of AI and technology to their advantage.

In recent years, there has been an increase in cyber attacks, which are becoming more sophisticated, frequent, and difficult to prevent. However, while AI provides new opportunities to combat cybercrime, it also introduces new risks and challenges. This article investigates how artificial intelligence is altering the landscape of cybersecurity and cybercrime.

Using AI to fight against cyber attacks

Traditional approaches to cybersecurity are becoming less effective as cyber-attacks grow in number and complexity. However, with the rise of AI, new opportunities to address these challenges are emerging. AI algorithms are capable of analyzing large amounts of data and detecting patterns and anomalies that humans may miss, which is especially promising for threat detection in cybersecurity. AI can prevent or mitigate potential damage through real-time detection and response.

Authentication and access control, which have traditionally been vulnerable to attacks, are also being transformed by AI. With AI-powered biometric authentication systems^[1] such as facial recognition, voice recognition, and fingerprint scanning, passwords, and other authentication methods can be made more secure and convenient. These systems are also capable of adapting to changes in user behavior, adding an extra layer of security.

AI-powered malware: the new threat to computer systems

AI-powered malware is becoming increasingly sophisticated and difficult to detect in the field of cybersecurity. In 2019, for example, researchers discovered DeepLocker,^[2] an AI-powered malware developed by IBM that uses “adversarial AI” to avoid detection. DeepLocker is intended to remain dormant until a specific target, such as an individual or organization, is detected. It can deploy a payload to deliver a destructive attack once it has identified the target.

Another example of AI-powered malware is the 2017 fileless malware attack on a global bank. The malware, dubbed “Silent Starling,”^[3] used an artificial intelligence algorithm to avoid detection by antivirus software. For several months, the malware was able to infect the bank's systems and steal confidential information without being detected. The attack was only discovered after an anomaly in the bank's network was discovered and traced back to the malware.

AI-powered malware attacks are especially concerning because they can adapt and evolve in real-time to avoid detection. These attacks can be used to steal sensitive data, hijack systems, or launch denial-of-service attacks,^[4] among other things. As AI technology advances, the threat of AI-powered malware is likely to grow, necessitating the development of new detection and prevention strategies.

The next generation of social engineering

ChatGPT, in the wrong hands, can assist anyone in creating a convincing phishing email and even writing the code for a malware attack.^[5] Unfortunately, regardless of your skill level or language, cybercriminals can now embed their newly created malicious code into an innocent-looking email attachment.

In recent years, there have been reports of cybercriminals carrying out social engineering attacks using similar tactics with other conversational AI models. Researchers, for example, discovered a phishing campaign in 2020 that used OpenAI's GPT-3 language model to generate convincing phishing emails that appeared to come from trusted sources.

It should be noted that using ChatGPT for social engineering attacks would be both unethical and illegal. OpenAI, the company that created ChatGPT, has put in place safeguards to prevent malicious use of its technology, such as monitoring for misuse and requiring users to agree to terms and conditions that prohibit harmful behavior. Nonetheless, the threat of AI-powered social engineering attacks is growing, emphasizing the importance of ongoing research and development of security measures to counter such threats.

Opportunities and risks to be addressed

AI provides new opportunities to combat cybercrime, but it also introduces new risks and challenges. AI-powered cybersecurity systems development is a promising area of research, but it is still in its early stages. Many technical and ethical issues must be addressed, such as the dependability and interpretability of AI algorithms, the possibility of bias and discrimination, and the risks of cyber attacks on AI systems themselves.

Another issue is the requirement for collaboration among cybersecurity professionals, AI experts, and policymakers. AI is a rapidly evolving technology, and developing effective solutions requires a multidisciplinary approach. Collaboration across disciplines is critical to ensuring that AI is used for the benefit of society rather than for malicious purposes.