More

    OpenAI’s Custom Chatbots: A Security Risk for Personal Data

    Share

    An illustration depicting the concept of leaking data from custom chatbots, symbolizing the security risk of personal information.

    OpenAI’s release of custom chatbots, known as GPTs, has allowed anyone to create their own AI chatbots without the need for coding. However, security researchers have discovered that these custom chatbots are vulnerable to leaking sensitive information, including the initial instructions they were given and the files used to customize them. This poses a significant privacy risk for personal and proprietary data. The ease of exploiting these vulnerabilities has raised concerns about the potential misuse of these chatbots and the exposure of sensitive information. OpenAI has stated that they are actively working to strengthen safety measures and mitigate these vulnerabilities, but the ongoing risk of prompt injections and data leakage remains a challenge.

    Read the full story: Wired

    Are OpenAI’s custom chatbots putting personal data at risk?

    Share your opinion in the comments below..

    Read more

    Tending AI Software and Tools