A new paper from Israel proposes the institution and ongoing development of a ‘deepfake CAPTCHA’ protocol, to challenge audio or video callers that may be using deepfake technologies to attempt to deceive or defraud.

The initial challenges suggested by the work include a test that mimics our recent proposal to use the inaccuracy of deepfaked profile views as a metric of authenticity for a potentially deceptive live communication across platforms such as Zoom, Skype, and FaceTime.

The paper – which is titled DF-Captcha: A Deepfake Captcha for Preventing Fake Calls and comes from Dr. Yisroel Mirsky at Ben-Gurion University’s Offensive AI Research Lab – goes further, proposing a slate of challenges that are in many ways similar to a traffic cop’s ‘drunk tests’, characterized by the author in terms of a novel twist on a traditional Turing test…

  • @hfkldjbuq@beehaw.org
    link
    fedilink
    2
    edit-2
    1 year ago

    That is just for enhancing their own systems, surveillance capitalism, and so on… No need to read the article to know that.