Deputy Editor
A 25-year-old wanted to jump on the ever-so-profitable ransomware bandwagon with the help of ChatGPT-like bots. Instead of making money, he received a prison sentence.
Tokyo District Court issued a three-year imprisonment sentence to Ryuki Hayashi for developing malware with the help of generative artificial intelligence (genAI), Japan Today reported.
GenAI chatbots, such as Antrhopic’s Claude, OpenAI’s ChatGPT, and other popular large language models (LLMs), can assist programmers in writing usable lines of code, a feature that didn’t go unnoticed by cybercriminals.
While genAI platforms use various methods to prevent attackers from crafting malicious code via their services, attackers find workarounds by rephrasing prompts that chatbots no longer deem dangerous.
Hayashi did not hide his true intentions, telling Japan’s media he “wanted to make money through ransomware” after he was arrested in May 2024. Ransomware operators encrypt data in target systems and demand money from the victims, threatening to delete or publicize their data.
According to Hayashi, he created the malicious code in six hours, perfectly illustrating how artificial intelligence lowers the entry bar for those willing to dabble in the cyber dark arts. Cybernews has discussed how attackers leverage AI for attacks, with industry insiders lamenting that, at least for now, cybercriminals benefit from genAI a lot more than defenders.
However, Hayashi’s sentencing, touted by Japan’s media as the first of its kind, could provide a valuable precedent. Even though Hayashi’s three-year prison sentence is suspended for four years, the country’s legal system shows little tolerance for AI-enabled hacking.
Tokyo District Court Judge Takashi Kawase reportedly said that Hayashi’s motives were clearly selfish, leaving no room for leniency. Thus, he demanded a prison sentence, not just a slap on the wrist. The sentence was suspended because the attacker admitted to committing the crime and showed remorse over his actions.