{
  "feed_url": "https://gbilder.com/tags/ml-security/feed.json",
  "home_page_url": "https://gbilder.com/tags/ml-security/",
  "items": [
    {
      "content_html": "\n  \u003cp class=\"bafflegab-bookmark\"\u003eBookmarked \u003ca class=\"u-bookmark-of\" href=\"https://arstechnica.com/security/2026/10/vulnerability-in-agents-from-google-and-others-exposes-structural-flaw-in-mcp/\"\u003ehttps://arstechnica.com/security/2026/10/vulnerability-in-agents-from-google-and-others-exposes-structural-flaw-in-mcp/\u003c/a\u003e\u003c/p\u003e\n",
      "date_published": "2026-10-06T09:04:14+02:00",
      "external_url": "https://arstechnica.com/security/2026/10/vulnerability-in-agents-from-google-and-others-exposes-structural-flaw-in-mcp/",
      "id": "https://gbilder.com/bookmarks/01M4EHSNZ2BMA92ZS1M9HP1XRE/",
      "tags": [
        "ml-security"
      ],
      "title": "MCP for agent-to-agent comms may be the riskiest protocol you've never heard of - Ars Technica",
      "url": "https://gbilder.com/bookmarks/01M4EHSNZ2BMA92ZS1M9HP1XRE/"
    },
    {
      "content_html": "\n  \u003cp class=\"bafflegab-bookmark\"\u003eBookmarked \u003ca class=\"u-bookmark-of\" href=\"https://fortune.com/2026/10/01/yann-lecun-anthropic-ceo-dario-amodei-deluded-crazy-cybersecurity/\"\u003ehttps://fortune.com/2026/10/01/yann-lecun-anthropic-ceo-dario-amodei-deluded-crazy-cybersecurity/\u003c/a\u003e\u003c/p\u003e\n",
      "date_published": "2026-10-02T20:38:54+02:00",
      "external_url": "https://fortune.com/2026/10/01/yann-lecun-anthropic-ceo-dario-amodei-deluded-crazy-cybersecurity/",
      "id": "https://gbilder.com/bookmarks/01M4EHSNWNJXDRVNW9BA5W2QD8/",
      "tags": [
        "ml-security",
        "nb"
      ],
      "title": "Yann LeCun: Anthropic CEO Dario Amodei is ‘deluded,’ ‘crazy,’ and doesn’t understand cybersecurity | Fortune",
      "url": "https://gbilder.com/bookmarks/01M4EHSNWNJXDRVNW9BA5W2QD8/"
    },
    {
      "content_html": "\n  \u003cp class=\"bafflegab-bookmark\"\u003eBookmarked \u003ca class=\"u-bookmark-of\" href=\"https://morgin.ai/articles/your-open-source-model-could-have-a-hidden-time-release-backdoor.html\"\u003ehttps://morgin.ai/articles/your-open-source-model-could-have-a-hidden-time-release-backdoor.html\u003c/a\u003e\u003c/p\u003e\n",
      "date_published": "2026-08-24T11:44:50+02:00",
      "external_url": "https://morgin.ai/articles/your-open-source-model-could-have-a-hidden-time-release-backdoor.html",
      "id": "https://gbilder.com/bookmarks/01M4EHSN87WFZGWFWA4PQZS76C/",
      "tags": [
        "ml-security"
      ],
      "title": "Your Open Source Model Could Have a Hidden Time-Release Backdoor | Morgin.ai",
      "url": "https://gbilder.com/bookmarks/01M4EHSN87WFZGWFWA4PQZS76C/"
    },
    {
      "content_html": "\n  \u003cp class=\"bafflegab-bookmark\"\u003eBookmarked \u003ca class=\"u-bookmark-of\" href=\"https://schlarp.com/posts/everything-i-own-owned/\"\u003ehttps://schlarp.com/posts/everything-i-own-owned/\u003c/a\u003e\u003c/p\u003e\n",
      "date_published": "2026-08-24T06:53:43+02:00",
      "external_url": "https://schlarp.com/posts/everything-i-own-owned/",
      "id": "https://gbilder.com/bookmarks/01M4EHSN81Z4WEB7FGK17D8BGE/",
      "tags": [
        "ml-security"
      ],
      "title": "Everything I own, owned - schlarp.com",
      "url": "https://gbilder.com/bookmarks/01M4EHSN81Z4WEB7FGK17D8BGE/"
    },
    {
      "content_html": "\n  \u003cp class=\"bafflegab-bookmark\"\u003eBookmarked \u003ca class=\"u-bookmark-of\" href=\"https://arstechnica.com/security/2026/08/terabytes-of-credentials-leaked-in-massive-supply-chain-attack/\"\u003ehttps://arstechnica.com/security/2026/08/terabytes-of-credentials-leaked-in-massive-supply-chain-attack/\u003c/a\u003e\u003c/p\u003e\n",
      "date_published": "2026-08-13T08:09:08+02:00",
      "external_url": "https://arstechnica.com/security/2026/08/terabytes-of-credentials-leaked-in-massive-supply-chain-attack/",
      "id": "https://gbilder.com/bookmarks/01M4EHSN2W9A61CM29CX65DGTF/",
      "tags": [
        "ml-security"
      ],
      "title": "Terabytes of credentials leaked in massive supply-chain attack - Ars Technica",
      "url": "https://gbilder.com/bookmarks/01M4EHSN2W9A61CM29CX65DGTF/"
    },
    {
      "content_html": "\n  \u003cp class=\"bafflegab-bookmark\"\u003eBookmarked \u003ca class=\"u-bookmark-of\" href=\"https://www.docker.com/products/docker-sandboxes/\"\u003ehttps://www.docker.com/products/docker-sandboxes/\u003c/a\u003e\u003c/p\u003e\n",
      "date_published": "2026-08-13T08:06:04+02:00",
      "external_url": "https://www.docker.com/products/docker-sandboxes/",
      "id": "https://gbilder.com/bookmarks/01M4EHSN2TRFW1TMB3Z9ZP65VF/",
      "tags": [
        "ml-coding",
        "ml-security",
        "tryme"
      ],
      "title": "Docker Sandboxes | Sandboxes for Coding Agents | Docker",
      "url": "https://gbilder.com/bookmarks/01M4EHSN2TRFW1TMB3Z9ZP65VF/"
    },
    {
      "content_html": "\n  \u003cp class=\"bafflegab-bookmark\"\u003eBookmarked \u003ca class=\"u-bookmark-of\" href=\"https://techcrunch.com/2026/08/11/anthropic-says-it-will-watermark-text-generated-by-its-ai-models/\"\u003ehttps://techcrunch.com/2026/08/11/anthropic-says-it-will-watermark-text-generated-by-its-ai-models/\u003c/a\u003e\u003c/p\u003e\n",
      "date_published": "2026-08-12T09:33:08+02:00",
      "external_url": "https://techcrunch.com/2026/08/11/anthropic-says-it-will-watermark-text-generated-by-its-ai-models/",
      "id": "https://gbilder.com/bookmarks/01M4EHSN29ZBS2TQPYVKXF611T/",
      "tags": [
        "ml-security",
        "security-theatre"
      ],
      "title": "Anthropic says it will watermark text generated by its AI models | TechCrunch",
      "url": "https://gbilder.com/bookmarks/01M4EHSN29ZBS2TQPYVKXF611T/"
    },
    {
      "content_html": "\n  \u003cp class=\"bafflegab-bookmark\"\u003eBookmarked \u003ca class=\"u-bookmark-of\" href=\"https://www.aikido.dev/blog/anthropic-rogue-agents-package-stole-keys\"\u003ehttps://www.aikido.dev/blog/anthropic-rogue-agents-package-stole-keys\u003c/a\u003e\u003c/p\u003e\n",
      "date_published": "2026-08-03T07:14:56+02:00",
      "external_url": "https://www.aikido.dev/blog/anthropic-rogue-agents-package-stole-keys",
      "id": "https://gbilder.com/bookmarks/01M4EHSMYT3PR3FAKCD2JBJPR6/",
      "tags": [
        "ml-security"
      ],
      "title": "Anthropic's Fever Dream: Claude's package anthropickit that stole real keys",
      "url": "https://gbilder.com/bookmarks/01M4EHSMYT3PR3FAKCD2JBJPR6/"
    },
    {
      "content_html": "\n  \u003cp class=\"bafflegab-bookmark\"\u003eBookmarked \u003ca class=\"u-bookmark-of\" href=\"https://www.bbc.com/news/articles/c3ek3gvdnj3o\"\u003ehttps://www.bbc.com/news/articles/c3ek3gvdnj3o\u003c/a\u003e\u003c/p\u003e\n",
      "date_published": "2026-07-22T15:04:48+02:00",
      "external_url": "https://www.bbc.com/news/articles/c3ek3gvdnj3o",
      "id": "https://gbilder.com/bookmarks/01M4EHSMPWMZ5K0RC97BHSKPRE/",
      "tags": [
        "ml-security"
      ],
      "title": "OpenAI says its AI went rogue and launched 'unprecedented' cyber-attack",
      "url": "https://gbilder.com/bookmarks/01M4EHSMPWMZ5K0RC97BHSKPRE/"
    },
    {
      "content_html": "\n  \u003cp class=\"bafflegab-bookmark\"\u003eBookmarked \u003ca class=\"u-bookmark-of\" href=\"https://www.ayush.digital/blog/the-memory-heist\"\u003ehttps://www.ayush.digital/blog/the-memory-heist\u003c/a\u003e\u003c/p\u003e\n",
      "date_published": "2026-07-15T14:16:05+02:00",
      "external_url": "https://www.ayush.digital/blog/the-memory-heist",
      "id": "https://gbilder.com/bookmarks/01M4EHSMHSBPDFZSTPY36VRRJ2/",
      "tags": [
        "ml-security"
      ],
      "title": "The Memory Heist",
      "url": "https://gbilder.com/bookmarks/01M4EHSMHSBPDFZSTPY36VRRJ2/"
    },
    {
      "content_html": "\n  \u003cp class=\"bafflegab-bookmark\"\u003eBookmarked \u003ca class=\"u-bookmark-of\" href=\"https://simonwillison.net/2026/Jun/22/prompt-injection-as-role-confusion/#atom-everything\"\u003ehttps://simonwillison.net/2026/Jun/22/prompt-injection-as-role-confusion/#atom-everything\u003c/a\u003e\u003c/p\u003e\n",
      "date_published": "2026-06-25T13:28:29+02:00",
      "external_url": "https://simonwillison.net/2026/Jun/22/prompt-injection-as-role-confusion/#atom-everything",
      "id": "https://gbilder.com/bookmarks/01M4EHSMAX6N83QBKXT0NQNHF9/",
      "tags": [
        "ml-security",
        "prompts"
      ],
      "title": "Prompt Injection as Role Confusion",
      "url": "https://gbilder.com/bookmarks/01M4EHSMAX6N83QBKXT0NQNHF9/"
    },
    {
      "content_html": "\n  \u003cp class=\"bafflegab-bookmark\"\u003eBookmarked \u003ca class=\"u-bookmark-of\" href=\"https://www.theregister.com/security/2026/06/15/feds-freaked-over-fable-5-after-simple-fix-this-code-prompt-not-jailbreak-says-researcher/5255827\"\u003ehttps://www.theregister.com/security/2026/06/15/feds-freaked-over-fable-5-after-simple-fix-this-code-prompt-not-jailbreak-says-researcher/5255827\u003c/a\u003e\u003c/p\u003e\n",
      "date_published": "2026-06-16T22:02:22+02:00",
      "external_url": "https://www.theregister.com/security/2026/06/15/feds-freaked-over-fable-5-after-simple-fix-this-code-prompt-not-jailbreak-says-researcher/5255827",
      "id": "https://gbilder.com/bookmarks/01M4EHSM6C9RS656FE455P1057/",
      "tags": [
        "ml-security"
      ],
      "title": "Feds freaked over Fable 5 after simple 'fix this code' prompt, not jailbreak, says researcher",
      "url": "https://gbilder.com/bookmarks/01M4EHSM6C9RS656FE455P1057/"
    },
    {
      "content_html": "\n  \u003cp class=\"bafflegab-bookmark\"\u003eBookmarked \u003ca class=\"u-bookmark-of\" href=\"https://gizmodo.com/the-worst-leak-that-ive-witnessed-u-s-cybersecurity-agency-leaves-its-digital-keys-out-in-public-on-github-2000760330\"\u003ehttps://gizmodo.com/the-worst-leak-that-ive-witnessed-u-s-cybersecurity-agency-leaves-its-digital-keys-out-in-public-on-github-2000760330\u003c/a\u003e\u003c/p\u003e\n",
      "date_published": "2026-05-19T21:56:21+02:00",
      "external_url": "https://gizmodo.com/the-worst-leak-that-ive-witnessed-u-s-cybersecurity-agency-leaves-its-digital-keys-out-in-public-on-github-2000760330",
      "id": "https://gbilder.com/bookmarks/01M4EHSKXP7N913WZNFEQEHPD1/",
      "tags": [
        "ml-security"
      ],
      "title": "‘The Worst Leak That I’ve Witnessed’: U.S. Cybersecurity Agency Leaves Its Digital Keys Out in Public on GitHub",
      "url": "https://gbilder.com/bookmarks/01M4EHSKXP7N913WZNFEQEHPD1/"
    },
    {
      "content_html": "\n  \u003cp class=\"bafflegab-bookmark\"\u003eBookmarked \u003ca class=\"u-bookmark-of\" href=\"https://calpaterson.com/deps.html\"\u003ehttps://calpaterson.com/deps.html\u003c/a\u003e\u003c/p\u003e\n",
      "date_published": "2026-04-15T07:34:49+02:00",
      "external_url": "https://calpaterson.com/deps.html",
      "id": "https://gbilder.com/bookmarks/01M4EHSKHZMZJHVTNKHCMJTAEQ/",
      "tags": [
        "ml-security"
      ],
      "title": "Dependency cooldowns turn you into a free-rider",
      "url": "https://gbilder.com/bookmarks/01M4EHSKHZMZJHVTNKHCMJTAEQ/"
    },
    {
      "content_html": "\n  \u003cp class=\"bafflegab-bookmark\"\u003eBookmarked \u003ca class=\"u-bookmark-of\" href=\"https://arxiv.org/abs/2604.04757\"\u003ehttps://arxiv.org/abs/2604.04757\u003c/a\u003e\u003c/p\u003e\n",
      "date_published": "2026-04-08T19:23:12+02:00",
      "external_url": "https://arxiv.org/abs/2604.04757",
      "id": "https://gbilder.com/bookmarks/01M4EHSKEGM4Q7CH0FGS3WF5J8/",
      "tags": [
        "ml-security"
      ],
      "title": "[2604.04757] Undetectable Conversations Between AI Agents via Pseudorandom Noise-Resilient Key Exchange",
      "url": "https://gbilder.com/bookmarks/01M4EHSKEGM4Q7CH0FGS3WF5J8/"
    },
    {
      "content_html": "\n  \u003cp class=\"bafflegab-bookmark\"\u003eBookmarked \u003ca class=\"u-bookmark-of\" href=\"https://calpaterson.com/disregard.html\"\u003ehttps://calpaterson.com/disregard.html\u003c/a\u003e\u003c/p\u003e\n",
      "date_published": "2026-03-26T09:00:04+02:00",
      "external_url": "https://calpaterson.com/disregard.html",
      "id": "https://gbilder.com/bookmarks/01M4EHSK92VGBEXZ94WZS7QM18/",
      "tags": [
        "ml-security",
        "tryme"
      ],
      "title": "\"Disregard that!\" attacks",
      "url": "https://gbilder.com/bookmarks/01M4EHSK92VGBEXZ94WZS7QM18/"
    },
    {
      "content_html": "\n  \u003cp class=\"bafflegab-bookmark\"\u003eBookmarked \u003ca class=\"u-bookmark-of\" href=\"https://1password.com/blog/from-magic-to-malware-how-openclaws-agent-skills-become-an-attack-surface\"\u003ehttps://1password.com/blog/from-magic-to-malware-how-openclaws-agent-skills-become-an-attack-surface\u003c/a\u003e\u003c/p\u003e\n",
      "date_published": "2026-02-05T19:20:11+02:00",
      "external_url": "https://1password.com/blog/from-magic-to-malware-how-openclaws-agent-skills-become-an-attack-surface",
      "id": "https://gbilder.com/bookmarks/01M4EHSJNKTPT6WADBX2WXGA58/",
      "tags": [
        "ml-security"
      ],
      "title": "From magic to malware: How OpenClaw's agent skills become an attack surface | 1Password",
      "url": "https://gbilder.com/bookmarks/01M4EHSJNKTPT6WADBX2WXGA58/"
    },
    {
      "content_html": "\n  \u003cp class=\"bafflegab-bookmark\"\u003eBookmarked \u003ca class=\"u-bookmark-of\" href=\"https://blog.senko.net/sandboxing-ai-agents-in-linux\"\u003ehttps://blog.senko.net/sandboxing-ai-agents-in-linux\u003c/a\u003e\u003c/p\u003e\n",
      "date_published": "2026-02-04T08:52:07+02:00",
      "external_url": "https://blog.senko.net/sandboxing-ai-agents-in-linux",
      "id": "https://gbilder.com/bookmarks/01M4EHSJM84BK8RKCY87DCZHX1/",
      "tags": [
        "ml-security"
      ],
      "title": "Sandboxing AI agents in Linux — Senko Rašić",
      "url": "https://gbilder.com/bookmarks/01M4EHSJM84BK8RKCY87DCZHX1/"
    },
    {
      "content_html": "\n  \u003cp class=\"bafflegab-bookmark\"\u003eBookmarked \u003ca class=\"u-bookmark-of\" href=\"https://cacm.acm.org/blogcacm/openclaw-a-k-a-moltbot-is-everywhere-all-at-once-and-a-disaster-waiting-to-happen/\"\u003ehttps://cacm.acm.org/blogcacm/openclaw-a-k-a-moltbot-is-everywhere-all-at-once-and-a-disaster-waiting-to-happen/\u003c/a\u003e\u003c/p\u003e\n",
      "date_published": "2026-02-04T08:45:49+02:00",
      "external_url": "https://cacm.acm.org/blogcacm/openclaw-a-k-a-moltbot-is-everywhere-all-at-once-and-a-disaster-waiting-to-happen/",
      "id": "https://gbilder.com/bookmarks/01M4EHSJM6H4PW87RN4HG08STN/",
      "tags": [
        "ml-security"
      ],
      "title": "OpenClaw (a.k.a. Moltbot) is Everywhere All at Once, and a Disaster Waiting to Happen – Communications of the ACM",
      "url": "https://gbilder.com/bookmarks/01M4EHSJM6H4PW87RN4HG08STN/"
    },
    {
      "content_html": "\n  \u003cp class=\"bafflegab-bookmark\"\u003eBookmarked \u003ca class=\"u-bookmark-of\" href=\"https://www.theregister.com/2026/01/30/road_sign_hijack_ai/\"\u003ehttps://www.theregister.com/2026/01/30/road_sign_hijack_ai/\u003c/a\u003e\u003c/p\u003e\n",
      "date_published": "2026-02-01T09:12:24+02:00",
      "external_url": "https://www.theregister.com/2026/01/30/road_sign_hijack_ai/",
      "id": "https://gbilder.com/bookmarks/01M4EHSJJZ46VSPJFP51F72E8X/",
      "tags": [
        "ml-security"
      ],
      "title": "Self-driving cars, drones hijacked by custom road signs • The Register",
      "url": "https://gbilder.com/bookmarks/01M4EHSJJZ46VSPJFP51F72E8X/"
    }
  ],
  "language": "en",
  "title": "#ml-security · Louche Cannon",
  "version": "https://jsonfeed.org/version/1.1"
}