AI companies leak data to advertisers [pdf]

(jorgegarciaherrero.com)

65 points | by damaru2 1 hour ago

9 comments

  • Coeur 37 minutes ago
    "multiple providers disclose sensitive conversation-derived artifacts — including titles, prompts, and screenshots — to third parties, often alongside persistent user identifiers that enable user attribution. We also find that some providers publicly expose conversation permalinks without access controls, allowing trackers to read the entire conversation."

    Not good at all.

    • postalcoder 4 minutes ago
      My least favorite trend I’ve noticed with so many AI chat services is they seem to equate a UUID url with privacy.

      Perplexity does this. Visiting a past perplexity search url exposes your full conversation.

  • gagan2020 8 minutes ago
    All sells but I saw Chinese models are upfront about that most of the time.
    • thenthenthen 0 minutes ago
      Also Chinese apps are all connected via various ad networks it seems even among competitors? I order something on a Tencent owned app and then I get ads in an Alibaba owned app for the same/similar product.
  • reedf1 6 minutes ago
    my first guess is always Gboard.
  • DrMandalay 14 minutes ago
    The word is "sell" not "leak". This title takes away all agency from the thieves selling private data to advertisers.
  • charcircuit 28 minutes ago
    The paper doesn't say when the app sends the conversion artifact.
  • classified 31 minutes ago
    Is it still called a leak if it was the whole point and purpose of the deal?

    Someone should have to investigate, but I suppose it's all "legal"?

    • lava_pidgeon 2 minutes ago
      In the US.

      In EU law it is very likely against GDPR.

  • folkrav 31 minutes ago
    Insert surprised pikachu meme
  • damaru2 1 hour ago
    Entire conversations via exposed permalinks. For Grok: trackers receiving the conversation URL could access the full chat because the link lacked access controls.

    Screenshots of conversations. TikTok received screenshots of Grok chats during sharing, exposing the actual visible conversation content.

    Conversation-derived content tied to persistent identifiers, including prompts and automatically generated chat titles revealing sensitive facts. "Salary 85k NYC: mortgage 280–350k".

  • irregularbowels 1 hour ago
    [dead]