• Lyrl@lemmy.dbzer0.com
    link
    fedilink
    English
    arrow-up
    8
    ·
    edit-2
    10 days ago

    Clicking through to the 404 Media article, I found this gem of a quote:

    In a March blog post called “Boosting Your Support and Safety on Meta’s Apps With AI” announcing its AI support feature, Meta said that the system can “Prevent an account takeover by noticing it was suddenly accessed from a new location, the password was changed, and edits were made to the profile—changes that, in isolation, look harmless to a person reviewing the account, but AI was able to recognize as a threat.”

    The very thing they boasted about the AI protecting against, not only did it not work, but it enabled that kind of attack. And they didn’t detect the exploit internally: this has been trending on Telegram since March, and only when the social media activity got large enough did they realize. Epic fail on Meta’s part.

    • GenosseFlosse@feddit.org
      link
      fedilink
      English
      arrow-up
      1
      ·
      9 days ago

      Meta said that the system can “Prevent an account takeover by noticing it was suddenly accessed from a new location, the password was changed, and edits were made to the profile—changes that, in isolation, look harmless to a person reviewing the account, but AI was able to recognize as a threat.”

      Explain how a computer cannot detect this much more efficiently without the use of ai, by checking the timestamps in some change logs?