[AI Revolution](68) Security Threats Growing with AI, Blocked by AI

The emergence of generative artificial intelligence (AI) has significantly improved work efficiency. While it is seen as a tool for creating new opportunities across various industries, it poses a major threat to security. This is because even ordinary people with no knowledge of hacking can now easily obtain hacking tools through generative AI. AhnLab, a first-generation domestic security company, has turned to AI to counter the increased security threats posed by AI.

AhnLab took its first step into the security industry in 1995 as a company specializing in antivirus software development. It has grown into a leading Korean information security company with the ‘V3’ product line, the longest-running software brand and antivirus solution in Korea. Operating 24/7 to prevent and block malware and hacking in real time, AhnLab has chosen to advance its solutions through AI to combat the new cyber security threats posed by AI.

Currently, AhnLab operates an ‘AI Team’ under its research institute that studies artificial intelligence and machine learning and develops related technologies. Since establishing its first dedicated AI department in 2018, it has continuously developed various detection models and researched foundational technologies.

Notably, AhnLab has applied ‘machine learning’ technology?AI research aimed at realizing functions similar to human learning ability on computers?to its solutions. Security solutions integrated with machine learning independently infer the likelihood of maliciousness for each threat element. They also provide evidence and synthesize it to reach a conclusion on whether something is malicious. Handling hundreds of thousands of cybersecurity threats daily is beyond human capacity; AI replaces humans in this task.

AhnLab’s ‘AhnLab XDR,’ the first domestically released ‘Extended Detection and Response (XDR)’ platform, also utilizes machine learning technology. It learns user and device behavior patterns to detect anomalies. For example, if an employee who works in Pangyo from 9 a.m. to 6 p.m. suddenly logs in from another country at 2 a.m. and sends an email, this would be detected as anomalous behavior.

Additionally, various AI technologies are embedded in malware detection, phishing email detection, smishing and phishing text message detection, and URL detection. Phishing refers to disguising as a trusted person or company to steal personal information. Smishing, a combination of SMS and phishing, involves clicking on URLs within text messages such as ‘mobile wedding invitations,’ which then install malware.

Appearance of a phishing email impersonating Apple Support Team detected by AhnLab.

Phishing emails, in particular, have become more sophisticated since the advent of generative AI. In the past, they were relatively easy to detect because they blatantly demanded money or used awkward sentences. However, recently, messages have become more refined using generative AI, making it important to understand the context of the message to determine whether it is phishing. AhnLab uses AI technology to extract and analyze comprehensive contextual information such as the sender’s address and content of phishing emails.

For example, AhnLab recently detected a phishing email disguised as the online streaming service ‘Netflix.’ The phishing email’s subject was ‘Your subscription is about to expire.’ Inside the email was a malicious URL containing the word ‘netfix,’ which is similar to ‘netflix.’ The AI-based phishing email detection model detected this email and blocked it from reaching users.

Going forward, AhnLab plans to advance machine learning technology for purposes such as ▲enhancing detection performance through anomaly-based threat detection and attack possibility prediction ▲explaining detection evidence and inferring entire attack scenarios ▲threat search and summarization, and apply these advancements to its solutions and services.

This content was produced with the assistance of AI translation services.