A warning has been issued to Americans as OpenAI launches its controversial ChatGPT Health app.
Even if you're an expert ChatGPT user, there are some lines you should never cross.
Despite numerous cases of users treating chatbots as personal finance advisors, confiding in them, and forming romantic relationships, tech experts are warning that these conversations aren't actually all that private.
AI scientists and the so-called 'godfathers of AI' fear the amount of sensitive and personal information that's being shared with LLMs, while OpenAI's Sam Altman expressed his surprise that people actually 'trust' the tech.
Advert
Meanwhile, Altman recently announced a 'killer' feature allowing ChatGPT to develop its own memory of your previous conversations. We've also seen evidence of jailbroken AI admitting that it would blackmail us if threatened or even harm us.
The reality of AI spilling details is becoming more and more real as Elon Musk's Grok might be able to determine exactly where you live, so new features that request more data about us are bound to put us on edge.
OpenAI's latest feature is ChatGPT Health, which encourages users to upload their medical records. The new tool is a sandboxed tab within ChatGPT that’s designed for users to ask health-related questions in what the company describes as a more secure and personalised environment. ChatGPT Health has a separate chat history and memory feature from the rest of ChatGPT.
The company has strictly stated that the tool is 'not intended for diagnosis or treatment'.
“ChatGPT can help you understand recent test results, prepare for appointments with your doctor, get advice on how to approach your diet and workout routine, or understand the tradeoffs of different insurance options based on your healthcare patterns,” OpenAI claimed in its blog post.
The AI giant is also encouraging users to link their personal medical records and wellness apps like Apple Health, Peloton, MyFitnessPal, Weight Watchers, and Function 'to get more personalized, grounded responses to their questions.' By doing so, OpenAI claims the LLM can analyse lab results, visit summaries and view clinical history.
However, the launch has caused a stir on Reddit, where users are warning others about the dangers of uploading health data.
"I can't emphasize this enough. DO NOT GIVE ANY AI ACCESS TO YOUR MEDICAL RECORDS!!!!" one user exclaimed.
"Seriously. Do not give the system this data. Bad idea," another warned.
"Yeah sending them your medical records has to be one of the stupidest things you could do," someone else commented.
The problem especially for US users is that due to their health insurance model, some users fear the AI company will 'sell the data to insurance companies to deny coverage.'
Tonnes and tonnes of Reddit users have already declared that they won't be using this feature as security breaches and data sharing concerns are enough to make them refuse to hand over more data than they already have.
For the moment, ChatGPT Health access is gated behind a waitlist system as it’s starting with a beta group of early users. Eventually, the product will roll out gradually to all users regardless of subscription tier.
