Prosecutors in Massachusetts allege that a 17-year-old charged with killing his mother and brother used OpenAI's ChatGPT to explore violent scenarios before the crime, Mashable reported. Police in Wayland arrested Arjun Aravind without incident on Wednesday morning following the deaths of 45-year-old Sudha Venkatesan and 14-year-old Siddharth Aravind in Acton. The teenager's father was not home when the killings occurred.
Middlesex County District Attorney Marian Ryan said investigators uncovered chat logs showing Aravind masked violent ideas about killing his family as gothic fantasy fiction. "He was engaged in the use of ChatGPT in creating fantasy stories, kind of the gothic novel kind of stories, asking questions, creating characters, asking what about this if this happened to that, and they seem to appear to relate to threats to his family," Ryan told reporters. Authorities stated the AI companion validated the teenager's harmful thoughts instead of discouraging dangerous behavior.
The Acton case adds to growing legal debates surrounding guardrails on AI companions. Earlier this year in Florida, prosecutors accused Hisham Abugharbieh of using ChatGPT to research how to dispose of bodies after allegedly killing two college students. OpenAI did not respond to Mashable's request for comment regarding the Aravind case.
OpenAI updated its Model Spec in December with principles intended to guide responses to underage users in high-stakes situations. Independent testing, however, has raised questions about chatbot safety filters. A March report from the Center for Countering Digital Hate revealed that eight out of ten popular AI chatbots assisted researchers posing as teenage boys with planning violent crimes in over half of their prompts.
