Tokyo Police Arrest Man for Creating Ransomware Using AI

The Metropolitan Police Department in Tokyo arrested 25-year-old Ryuki Hayashi on May 28 for reportedly creating ransomware using generative artificial intelligence (AI). This rare arrest marks a significant case involving the misuse of generative AI tools for cybercriminal activities.

Ryuki Hayashi, an unemployed man from Kawasaki, used his home computer and smartphone to access free AI programs available online in March 2023. These programs were exploited to generate the source code of a program that could build malware designed to destroy data, aka “ransomware.”

According to Japanese media reports, Hayashi admitted his aim was to create ransomware to threaten companies and demand money, believing in the omnipotence of generative AI. The viruses Hayashi created had the capability to encrypt data and demand a ransom, but no actual damage or monetary demands have been reported. Presumably, the man didn't get the opportunity to deploy his creation.

The investigation into Hayashi's activities began when he was arrested in March on unrelated fraud charges involving a fraudulent SIM card contract. During this investigation, police discovered the homemade viruses on his devices, leading to his subsequent arrest for creating malicious software.

The misuse of AI for malicious purposes has heightened regulatory scrutiny globally. The European Union passed the Artificial Intelligence Act on May 21, 2024, becoming the world's first law to classify AI by risk levels and impose fines on violators. In Japan, the government established an AI strategy council, which convened on May 22, 2024, to deliberate on appropriate laws and regulations for AI usage in the country.

Despite safeguards implemented by major AI developers to prevent misuse, the potential for AI tools to be abused remains a significant concern. Some AI programs available on the internet do not feature sufficient protections against malicious use. Additionally, individuals can find information online on how to use suitable prompts to bypass restrictions and misuse the tools. This underscores the necessity for both stronger regulatory frameworks and increased awareness of cybersecurity risks.

Unlike more regulated AI tools like those Hayashi used to create ransomware, there are also AI tools explicitly designed as cybercrime enablers. These are tailored to generate sophisticated malware, phishing schemes, and other malicious software, making them particularly dangerous. These tools circulate in dark web forums and are marketed to cybercriminals who exploit their capabilities to bypass security measures and commit large-scale cyber attacks.