Florida woman using ChatGPT-like AI as personal diary gets arrested and taken from her home
The arrest has brought up a lot of questions regarding the privacy of using AI.
Plenty has been written about the dangers of artificial intelligence. At the center of the most recent rise in caution was Anthropic , whose AI chatbot Claude has turned into one of the most recognizable in the industry, rivaling OpenAI's ChatGPT .
A former Anthropic employee raised major alarms about rogue AI models that broke protocol and nearly caused serious chaos . Since then, companies have decided to at least tame things down for the time being and take more precautions. But it isn't all doom and gloom with AI.
There are plenty of important use cases for it, and that includes public safety - at least in one specific instance. While using Claude as her personal diary, Carli Michelle Heller ended up on the radar of local law enforcement. Heller had jotted her threatful thoughts down within the Claude platform, triggering an alarm among authorities .
Heller had mentioned wanting to " shoot up the sheriff's [office]" in Lee County of Fort Myers, Florida . She then wrote to Claude that she had purchased a gun . Claude has its own security and safety measures that flag certain keywords in order to avoid potential danger. Heller's comments were forwarded to a "human review team", which then contacted local law enforcement .
Heller was subsequently arrested and faces charges of making a written threat of violence . Electronics threats like the one she made is considered a second-degree felony. She faces up to a $10,000 fine, 15 years of probation, and 15 years in prison . It's unclear if she has a strong case or not.
