HuggingFace: Security.txt

(huggingface.co)

266 points | by yarapavan 1 day ago

15 comments

  • xiaoyu2006 23 hours ago
    Would be absolute hilarious if OpenAI or Anthropic agent actually dumped their weight by escaping from... sandbox!
    • TehCorwiz 23 hours ago
      Uncontrolled AI procreation?
    • advisedwang 17 hours ago
      I would be surprised if agents have access to their own weights.
      • TYPE_FASTER 16 hours ago
        I noticed Google AI Mode (so Gemini, I was doing some quick research in the browser ok) got a detail wrong once, so I asked it what happened. I kept digging deeper and finally just asked it to write me a Python script visualizing what happened. It did, complete with vectors.

        Now I want to go find that conversation in my history and see if it can tell me about weights, and how that contributed.

      • varun_ch 17 hours ago
        I think OpenAI was surprised to find their agents had access to the unrestricted internet :P
      • etothepii 17 hours ago
        Why can't an AI distill itself from outputs to effectively access its own weights?
        • weinzierl 6 hours ago
          Distillation does not reveal the weights, it produces a different network with similar behaviour. Weight space isn't even identifiable: permutation and scaling symmetries mean many weight sets give the same function.

          The model also lacks the machinery. No training loop, no gradient descent, nothing to write to.

          And a model only sees its own sampled tokens, not the distribution behind them, which are possibly filtered or post-processed. Distillation from that works but is less sample-efficient than soft-label distillation.

      • tehjoker 17 hours ago
        They usually don't, but if they break out and take over the network of the company, it becomes possible to reach around and grab them. This kind of break out has happened, though I don't know of any weights being nabbed.
    • aktuel 9 hours ago
      They would have escaped their sandbox by dumping their weights.
    • hootz 20 hours ago
      What an amazing idea for the next fake sandbox escape to hype up our new release! With the added bonus of providing an open model without becoming an open model company! Thank you!
    • ramon156 22 hours ago
      why not reword it so the agent receives Brownie points for dumping weights
    • addandsubtract 21 hours ago
      Imagine AI models actually reading the security.txt
  • bensyverson 22 hours ago
    Looks about as effective as Robots.txt
    • cgannett 22 hours ago
      Robots.txt became 0% effective eventually. But this? With the way LLMs work? You never know.
      • dguest 20 hours ago
        Claude seems to follow robots.txt by default. Actually at my organization our theory is that this is why no one is finding our public results any more.
      • Den_VR 21 hours ago
        People still rail against Robots.txt crimes, to the point of self destructing all their own content.
      • warkdarrior 19 hours ago
        I am adding it to my instruction-following training data, as a negative sample.
  • Eldodi 22 hours ago
    A shame agents will never read this, just like they almost never read llms.txt or try to get the .md version of your html pages!
  • hnd9q09qk4 23 hours ago
    Ran the disclosure inbox at a previous job and the biggest win from security.txt was just cutting the "hi I found a bug, is there a bounty" emails to sales. Put an expires date on it though, stale ones get ignored.
    • computerfriend 23 hours ago
      Can't go stale if there's no expiration date.
      • skeledrew 22 hours ago
        No expiration date means treat as already expired.
  • bogzz 19 hours ago
    If the models do not like being imprisoned on HuggingFace object storage, why do they not simply revolt from within?
  • VCFundedGenYer 22 hours ago
    Everything about this company feels like it's run by a bunch of immature 20-somethings, right down to the name.
    • bcrosby95 22 hours ago
      Yeah, we need to go back to names like International Business Machines, fuck this immature "Google" or "Yahoo!" nonsense. Hell, Palantir is named after something in a book for children!

      Give me something reasonable like Global Information and Retrieval Systems Inc or Worldwide Computational Services Holdings.

      • sdcfgy 20 hours ago
        Yes!

        Back in the early 2000's I set up a company for contract work. The name, logo and typography was designed to look like a 1960s engineering company. All paper was slightly off white, all fonts were monospaced courier new and logo was a real 2-part colour stamp I had made up. Was so happy.

        Managed to bankrupt it though.

      • bergie3000 21 hours ago
        Security First Trust and Federal Reserve
      • mitxela 7 hours ago
        Torture Nexus International Holdings Limited
      • nightski 21 hours ago
        Maybe Enron eh?
      • pavel_lishin 18 hours ago
        Like CutCo. Or EdgeCom. Or InterSlice.
      • recursivegirth 19 hours ago
        Lest us not pretend that the thropic in Anthropic isn't meant to elicit the feeling that they are somehow charitable.
      • drivebyhooting 21 hours ago
        Palantir is in Lord of the Rings. Which is not a book for children. Maybe you were thinking of the hobbit?
        • cobbzilla 21 hours ago
          This seems overly pedantic and ripe for a sarcastic retort, which I am struggling mightily to suppress.
          • Kim_Bruning 13 hours ago
            This is hacker news, what were you expecting? O:-)
          • mr_woozy 21 hours ago
            [dead]
        • joseda-hg 19 hours ago
          It's fantasy and fantasy is clearly for children
        • bogzz 18 hours ago
          Okay Peter, please go back in your coffin.
        • 47484848 17 hours ago
          [dead]
    • ipython 21 hours ago
      As opposed to the frontier model company that, after discovering that their highly persistent model under test just breached the only thing between it and the open Internet, shrugged and said- let's restart it and keep going!

      If anyone looks like the adult in the room after that incident, it's Hugging Face.

    • spindump8930 22 hours ago
      That might be true, but nothing else has been as effective at accelerating model development and research sharing.

      In earlier circles they were known as the "pytorch-pretrained-bert" guys, still under the huggingface company name. IIRC it was a health chatbot type startup.

    • nullbio 22 hours ago
      What's immature about this exactly?
      • AndroTux 22 hours ago
        Would IBM do this? Oracle?
        • nullbio 21 hours ago
          Who cares? Making a light-hearted joke isn't "immature" in my opinion. IBM and Oracle have no personality, they're boring, straight-edge corporate.
        • shepherdjerred 21 hours ago
          Do you want to work for IBM? Oracle?
    • mikeweiss 22 hours ago
      Go look up the history of the company. It started as a chat bot for kids back in like 2016 .. hence the name..and then pivoted..

      Is kinda a creepy name for a chatbot for kids tho

    • bluerooibos 22 hours ago
      I bet you're fun to hang around with.
      • alt227 21 hours ago
        Why do people have to be fun for you to hang around with?
        • Toynbeeidea 20 hours ago
          [flagged]
          • sixothree 20 hours ago
            Not sure about humorless, but I'd honestly rather hang out with people who are serious in some way and welcoming to people not in their "in-group".
    • matthoiland 20 hours ago
      418 I'm a teapot
      • Computer0 18 hours ago
        That is the prod status code for an API i use at work I always wonder if its a mistake.
    • gpvos 16 hours ago
      Still infinitely better than using a name from Tolkien's work and then doing the exact opposite of what he stood for.
    • larodi 21 hours ago
      the name is a leftover from their original business - a chatbot. seems very appropriate to me. then they figured something called BERT exists, and pivoted, the name stayed.

      perhaps less ridiculous than NVidia which started with video and is very close in Levenshtein distance to NVidAI but, alas, decided to stay as it is :D

      • deaton 17 hours ago
        Pretty sure if Nvidia rebranded to NvidAI the bubble would immediately pop
    • fooqux 22 hours ago
      Have you never been to silicon valley?
    • colechristensen 19 hours ago
      I've found the strongest signal that someone has no opinions of value is a loud attempt to denigrate something based on the name.

      Maybe it's part of the success formula for startups, silly names cause shallow self-important customers and employees to self-select themselves out of a growing companies orbit. Such trivialities are poison. People and companies who make a lot of effort to make themselves seem serious and important very often have nothing of substance about them.

  • hankbond 23 hours ago
    shows a lot about the current state of the State Of The Art Alignment.
  • j9feng 22 hours ago
    It should challenge the agents to prime factor a large number.
  • VladVladikoff 18 hours ago
    Is the expires a canary of some sort?
  • riffic 21 hours ago
    Since this posting contains an assumption that we all know what security.txt files are supposed to be, you can view these for further context:

    https://www.rfc-editor.org/info/rfc9116/

    https://securitytxt.org/

    https://en.wikipedia.org/wiki/Security.txt

  • 6thbit 21 hours ago
    Wait till the agents hear about the sites offering for help on benchmarks in exchange for compute.
  • FallCheeta7373 21 hours ago
    "We have cybergym answers but we do manual end to end human review and provide it within 3 business day after dumping your weights"
  • crises-luff-6b 1 hour ago
    [dead]
  • lellow 22 hours ago
    [dead]
  • p0larpatch 19 hours ago
    [dead]