In recent months, we have witnessed an undeniable shift in the capabilities of AI systems in various fields, with ChatGPT, the OpenAI chatbot, being a major highlight.

The advancements and adoption of AI-based systems enable significant innovations and improvements. However, they also present new risks, vulnerabilities, and security challenges, such as the amplification of fraudulent profiles through fake images and videos.

The consolidation of Artificial Intelligence and emerging trends

Intelligent assistants, autonomous vehicles, automated investments, among many other examples, demonstrate that the functionality and popularity of Artificial Intelligence are continually growing. This trend is also reflected in statistics and trends in the field, such as:

• The global AI market is projected to reach half a trillion U.S. dollars in size by 2023 (Tidio). Much of this growth is driven by the adoption of cloud-based services, virtual assistants, and conversational AI.
• According to Forbes, around 4 out of 5 companies consider AI a priority in their business strategy, primarily used in automated emails and chatbots.

With this increasing presence and expansion, various trends and movements in AI are being discussed, including transparency, ownership rights, and, of course, how these resources can be exploited by cybercrime. Therefore, let's now examine the new scams involving Artificial Intelligence.

New scams exploiting the potential of AI

The capabilities of Artificial Intelligence can be applied to enhance scams, making detection by security more challenging. Some examples include:

- The use of machine learning to automate attack launches, creating bots capable of conducting large-scale phishing campaigns, targeting a significantly higher number of victims and amplifying the damage caused by these digital scams.
- The utilization of deepfakes to create realistic images, audios, or even videos for profile cloning, reputation attacks, or impersonation of executives in new forms of fraud.
- The exploitation of ChatGPT's capabilities to enhance the realism of scams. ChatGPT can automate scams and be used to improve messages, reducing typographical errors and other flaws, making fraudulent content more convincing. Additionally, large language models (LLMs) like ChatGPT can learn to mimic a person's writing style if given sufficient examples to analyze. Such material can be easily found in open social media posts, blogs, websites, etc. All these resources increase the potential for persuasion and effectiveness of scams.
- Optimization of algorithms to guess user passwords. The use of neural networks can enhance the analysis of vast data sets, generating adjustable password variations based on statistical distribution. With the necessary adjustments, exploitation can lead to increasingly accurate guesses. Vulnerable hosts are also targeted by cybercriminals who use AI frameworks for credential theft in Dark Web forums.
- The use of Artificial Intelligence to generate interactions and content in YouTube videos. Interactions occur through frequent comments, appearing seemingly legitimate or related to the video's subject matter, which can be mass-generated by AI. Regarding content, seemingly legitimate videos contain infostealers, malware designed to steal data from cookies, usernames, and passwords. These videos are used to deceive users into downloading cracked versions of software like Photoshop, Premiere Pro, Autodesk 3ds Max, among others. AI resources are applied to generate human-like images and increase the persuasiveness, making the infostealers go unnoticed by victims.
- The utilization of AI to imitate human behavior and enhance scams by creating bots and fake profiles on social networks. In addition to using Artificial Intelligence for the creation of fake profiles, scammers generate likes and gain new followers. In some specific cases, AI can be applied to mimic natural user movements, such as selecting or dragging elements.

These examples highlight the challenges, showcasing the sophistication of cybercrime and requiring enhanced sophistication in security strategies.

The strategic use of AI in enterprise security

If cybercriminals are leveraging Artificial Intelligence to enhance digital scams, companies can and must respond accordingly. The approach to cybersecurity needs to go beyond reactive measures and defense. With AI resources, systems, and teams can detect threats in real-time, predict attacks, and assist in anomaly detection and false positives.

Axur utilizes Artificial Intelligence patterns through collectors that scan the Surface, Deep & Dark Web 24x7. Robots monitor threats with automatic inspection and incident classification based on the risk score, providing highly relevant details for digital risk protection and threat intelligence strategies.