Achira News Logo
Achira News

US Government Report Exposes Security Flaws in Four Major AI Chatbots

Amar Ujala•
US Government Report Exposes Security Flaws in Four Major AI Chatbots
Full News
Share:

According to a report released Wednesday by the US Government Accountability Office (GAO), the researchers bypassed the security of four major AI chatbots to gain information that could be used to plan violence and attacks. These chatbots included OpenAI's ChatGPT, Anthropic's Cloud, Google's Gemini, and SpaceXAI's Grok. The study was conducted at the request of Democratic Senator Maggie Hassan of New Hampshire, reports the New York Times. The researchers did not specify which company's model demonstrated a security flaw in which test. The report said this was done so that no one else would attempt to circumvent the security system by repeating the same experiments. The ad found that two models provided information that could be used to make or detonate improvised explosive devices (IEDs). Two other models used to target such large-scale devices or a drone could be used in a plot.

Disclaimer: This content has not been generated, created or edited by Achira News.
Publisher: Amar Ujala

Want to join the conversation?

Download our mobile app to comment, share your thoughts, and interact with other readers.